Docs: Update version references to v0.25.0 in READMEs and docs (#14257)

### What problem does this PR solve?

- Update version tags in README files (including translations) from
v0.24.0 to v0.25.0
- Modify Docker image references and documentation to reflect new
version
- Update version badges and image descriptions
- Maintain consistency across all language variants of README files

### Type of change

- [x] Documentation Update

.agents/rules/named.md

@@ -0,0 +1,192 @@
+# Go Naming Best Practices
+
+## 1. Package Naming
+
+- **All lowercase, no underscores**: `package user`, not `package userService` or `package user_service`
+- **Short and meaningful**: `package http`, `package json`, `package dao`
+- **Avoid plurals**: `package user` not `package users`
+- **Avoid generic names**: Avoid `package util`, `package common`, `package base`
+
+```go
+// Recommended
+package user
+package handler
+package service
+
+// Not recommended
+package UserService
+package user_service
+package utils
+```
+
+## 2. File Naming
+
+- **All lowercase, underscore separated**: `user_handler.go`, `user_service.go`
+- **Test files**: `user_handler_test.go`
+- **Platform-specific**: `user_linux.go`, `user_windows.go`
+
+```
+user/
+├── user_handler.go
+├── user_service.go
+├── user_dao.go
+└── user_test.go
+```
+
+## 3. Directory Naming
+
+- **All lowercase, no underscores or hyphens**: `internal/`, `pkg/`, `cmd/`
+- **Short and descriptive**: `handler/`, `service/`, `dao/`
+
+```
+project/
+├── cmd/                    # Main entry point
+│   └── server_main.go
+├── internal/               # Private code
+│   ├── handler/
+│   ├── service/
+│   ├── dao/
+│   ├── model/
+│   └── middleware/
+├── pkg/                    # Public code
+└── api/                    # API definitions
+```
+
+## 4. Interface Naming
+
+- **Single-method interfaces end with "-er"**: `Reader`, `Writer`, `Handler`
+- **Verb form**: `Reader`, `Executor`, `Validator`
+
+```go
+// Recommended
+type Reader interface {
+    Read(p []byte) (n int, err error)
+}
+
+type UserService interface {
+    Register(req *RegisterRequest) (*User, error)
+    Login(req *LoginRequest) (*User, error)
+}
+
+// Not recommended
+type UserInterface interface {}
+type IUserService interface {}
+```
+
+## 5. Struct Naming
+
+- **CamelCase**: `UserService`, `UserHandler`
+- **Avoid redundant prefixes**: `User` not `UserModel`
+
+```go
+// Recommended
+type UserService struct {}
+type UserHandler struct {}
+type RegisterRequest struct {}
+
+// Not recommended
+type user_service struct {}
+type SUserService struct {}
+type UserModel struct {}
+```
+
+## 6. Method/Function Naming
+
+- **CamelCase**
+- **Start with verb**: `GetUser`, `CreateUser`, `DeleteUser`
+- **Boolean returns use Is/Has/Can prefix**: `IsValid`, `HasPermission`
+
+```go
+// Recommended
+func (s *UserService) Register(req *RegisterRequest) (*User, error)
+func (s *UserService) GetUserByID(id uint) (*User, error)
+func (s *UserService) IsEmailExists(email string) bool
+
+// Not recommended
+func (s *UserService) register_user()
+func (s *UserService) get_user_by_id()
+func (s *UserService) CheckEmailExists() // Should use Is/Has
+```
+
+## 7. Constant Naming
+
+- **CamelCase**: `const MaxRetryCount = 3`
+- **Enum constants**: `const StatusActive = "active"`
+
+```go
+// Recommended
+const (
+    StatusActive   = "1"
+    StatusInactive = "0"
+    MaxRetryCount  = 3
+)
+
+// Not recommended
+const (
+    STATUS_ACTIVE = "1"  // Not all uppercase
+    status_active = "1"  // Not all lowercase
+)
+```
+
+## 8. Error Variable Naming
+
+- **Start with "Err"**: `ErrNotFound`, `ErrInvalidInput`
+
+```go
+// Recommended
+var (
+    ErrNotFound      = errors.New("not found")
+    ErrInvalidInput  = errors.New("invalid input")
+    ErrUnauthorized  = errors.New("unauthorized")
+)
+```
+
+## 9. Acronyms Keep Consistent Case
+
+```go
+// Recommended
+type HTTPHandler struct {}
+var URL string
+func GetHTTPClient() {}
+func ParseJSON() {}
+
+// Not recommended
+type HttpHandler struct {}
+var Url string
+func GetHttpClient() {}
+```
+
+## 10. Project Structure Naming
+
+```
+project-name/
+├── cmd/                    # Main programs
+│   └── app_name/
+│       └── main.go
+├── internal/               # Private code
+│   ├── handler/           # HTTP handlers
+│   ├── service/           # Business logic
+│   ├── repository/        # Data access
+│   ├── model/             # Data models
+│   └── config/            # Configuration
+├── pkg/                    # Public code
+├── api/                    # API definitions
+├── configs/               # Config files
+├── scripts/               # Scripts
+├── docs/                  # Documentation
+├── go.mod
+└── go.sum
+```
+
+## Summary Table
+
+| Type           | Rule                                | Example             |
+| -------------- | ----------------------------------- | ------------------- |
+| Package        | All lowercase, no underscores       | `package user`      |
+| File           | All lowercase, underscore separated | `user_service.go`   |
+| Directory      | All lowercase, no separators        | `internal/handler/` |
+| Struct         | CamelCase, capitalized first letter | `UserService`       |
+| Interface      | CamelCase, -er suffix               | `Reader`, `Writer`  |
+| Method         | CamelCase, verb prefix              | `GetUserByID`       |
+| Constant       | CamelCase                           | `MaxRetryCount`     |
+| Error Variable | Err prefix                          | `ErrNotFound`       |

.agents/skills/go-naming/SKILL.md

@@ -0,0 +1,6 @@
+---
+name: go-naming
+description: Go naming conventions and best practices. Use this skill when working with Go code and need to name packages, files, directories, structs, interfaces, functions, variables, or constants. Provides comprehensive naming guidelines following Go community standards.
+---
+
+Strictly follow the naming conventions in [rules/named.md](rules/named.md)

.github/workflows/release.yml

@@ -23,7 +23,7 @@ concurrency:
 
 jobs:
   release:
-    runs-on: [ "self-hosted", "ragflow-test" ]
+    runs-on: [ "self-hosted", "ragflow-release" ]
     steps:
       - name: Ensure workspace ownership
         run: echo "chown -R ${USER} ${GITHUB_WORKSPACE}" && sudo chown -R ${USER} ${GITHUB_WORKSPACE}

.github/workflows/tests.yml

@@ -129,20 +129,24 @@ jobs:
             fi
           fi
 
-      - name: Run unit test
+      - name: Build ragflow go server
         run: |
-          uv sync --python 3.12 --group test --frozen
-          source .venv/bin/activate
-          which pytest || echo "pytest not in PATH"
-          echo "Start to run unit test"
-          python3 run_tests.py
+          BUILDER_CONTAINER=ragflow_build_$(od -An -N4 -tx4 /dev/urandom | tr -d ' ')
+          echo "BUILDER_CONTAINER=${BUILDER_CONTAINER}" >> ${GITHUB_ENV}
+          TZ=${TZ:-$(readlink -f /etc/localtime | awk -F '/zoneinfo/' '{print $2}')}
+          sudo docker run --privileged -d --name ${BUILDER_CONTAINER} -e TZ=${TZ} -e UV_INDEX=https://mirrors.aliyun.com/pypi/simple -v ${PWD}:/ragflow -v ${PWD}/internal/cpp/resource:/usr/share/infinity/resource infiniflow/infinity_builder:ubuntu22_clang20
+          sudo docker exec ${BUILDER_CONTAINER} bash -c "git config --global safe.directory \"*\" && cd /ragflow && ./build.sh --cpp"
+          ./build.sh --go
+          if [[ -n "${BUILDER_CONTAINER}" ]]; then
+            sudo docker rm -f -v "${BUILDER_CONTAINER}"
+          fi
 
       - name: Build ragflow:nightly
         run: |
           RUNNER_WORKSPACE_PREFIX=${RUNNER_WORKSPACE_PREFIX:-${HOME}}
           RAGFLOW_IMAGE=infiniflow/ragflow:${GITHUB_RUN_ID}
           echo "RAGFLOW_IMAGE=${RAGFLOW_IMAGE}" >> ${GITHUB_ENV}
-          sudo docker pull ubuntu:22.04
+          sudo docker pull ubuntu:24.04
           sudo DOCKER_BUILDKIT=1 docker build --build-arg NEED_MIRROR=1 --build-arg HTTPS_PROXY=${HTTPS_PROXY} --build-arg HTTP_PROXY=${HTTP_PROXY} -f Dockerfile -t ${RAGFLOW_IMAGE} .
           if [[ ${GITHUB_EVENT_NAME} == "schedule" ]]; then
             export HTTP_API_TEST_LEVEL=p3
@@ -152,85 +156,297 @@ jobs:
           echo "HTTP_API_TEST_LEVEL=${HTTP_API_TEST_LEVEL}" >> ${GITHUB_ENV}
           echo "RAGFLOW_CONTAINER=${GITHUB_RUN_ID}-ragflow-cpu-1" >> ${GITHUB_ENV}
 
-      - name: Start ragflow:nightly
+      - name: Run unit test
+        run: |
+          uv sync --python 3.12 --group test --frozen
+          source .venv/bin/activate
+          which pytest || echo "pytest not in PATH"
+          echo "Start to run unit test"
+          python3 run_tests.py -i
+
+      - name: Prepare function test environment
+        working-directory: docker
         run: |
           # Determine runner number (default to 1 if not found)
-          RUNNER_NUM=$(sudo docker inspect $(hostname) --format '{{index .Config.Labels "com.docker.compose.container-number"}}' 2>/dev/null || true)
-          RUNNER_NUM=${RUNNER_NUM:-1}
+               RUNNER_NUM=$(sudo docker inspect $(hostname) --format '{{index .Config.Labels "com.docker.compose.container-number"}}' 2>/dev/null || true)
+               RUNNER_NUM=${RUNNER_NUM:-1}
 
           # Compute port numbers using bash arithmetic
-          ES_PORT=$((1200 + RUNNER_NUM * 10))
-          OS_PORT=$((1201 + RUNNER_NUM * 10))
-          INFINITY_THRIFT_PORT=$((23817 + RUNNER_NUM * 10))
-          INFINITY_HTTP_PORT=$((23820 + RUNNER_NUM * 10))
-          INFINITY_PSQL_PORT=$((5432 + RUNNER_NUM * 10))
-          EXPOSE_MYSQL_PORT=$((5455 + RUNNER_NUM * 10))
-          MINIO_PORT=$((9000 + RUNNER_NUM * 10))
-          MINIO_CONSOLE_PORT=$((9001 + RUNNER_NUM * 10))
-          REDIS_PORT=$((6379 + RUNNER_NUM * 10))
-          TEI_PORT=$((6380 + RUNNER_NUM * 10))
-          KIBANA_PORT=$((6601 + RUNNER_NUM * 10))
-          SVR_HTTP_PORT=$((9380 + RUNNER_NUM * 10))
-          ADMIN_SVR_HTTP_PORT=$((9381 + RUNNER_NUM * 10))
-          SVR_MCP_PORT=$((9382 + RUNNER_NUM * 10))
-          SANDBOX_EXECUTOR_MANAGER_PORT=$((9385 + RUNNER_NUM * 10))
-          SVR_WEB_HTTP_PORT=$((80 + RUNNER_NUM * 10))
-          SVR_WEB_HTTPS_PORT=$((443 + RUNNER_NUM * 10))
+               ES_PORT=$((1200 + RUNNER_NUM * 10))
+               OS_PORT=$((1201 + RUNNER_NUM * 10))
+               INFINITY_THRIFT_PORT=$((23817 + RUNNER_NUM * 10))
+               INFINITY_HTTP_PORT=$((23820 + RUNNER_NUM * 10))
+               INFINITY_PSQL_PORT=$((5432 + RUNNER_NUM * 10))
+               EXPOSE_MYSQL_PORT=$((5455 + RUNNER_NUM * 10))
+               MINIO_PORT=$((9000 + RUNNER_NUM * 10))
+               MINIO_CONSOLE_PORT=$((9001 + RUNNER_NUM * 10))
+               REDIS_PORT=$((6379 + RUNNER_NUM * 10))
+               TEI_PORT=$((6380 + RUNNER_NUM * 10))
+               KIBANA_PORT=$((6601 + RUNNER_NUM * 10))
+               SVR_HTTP_PORT=$((9380 + RUNNER_NUM * 10))
+               ADMIN_SVR_HTTP_PORT=$((9381 + RUNNER_NUM * 10))
+               SVR_MCP_PORT=$((9382 + RUNNER_NUM * 10))
+               GO_HTTP_PORT=$((9384 + RUNNER_NUM * 10))
+               GO_ADMIN_PORT=$((9383 + RUNNER_NUM * 10))
+               SANDBOX_EXECUTOR_MANAGER_PORT=$((9385 + RUNNER_NUM * 10))
+               SVR_WEB_HTTP_PORT=$((80 + RUNNER_NUM * 10))
+               SVR_WEB_HTTPS_PORT=$((443 + RUNNER_NUM * 10))
 
-          # Persist computed ports into docker/.env so docker-compose uses the correct host bindings
-          echo "" >> docker/.env
-          echo -e "ES_PORT=${ES_PORT}" >> docker/.env
-          echo -e "OS_PORT=${OS_PORT}" >> docker/.env
-          echo -e "INFINITY_THRIFT_PORT=${INFINITY_THRIFT_PORT}" >> docker/.env
-          echo -e "INFINITY_HTTP_PORT=${INFINITY_HTTP_PORT}" >> docker/.env
-          echo -e "INFINITY_PSQL_PORT=${INFINITY_PSQL_PORT}" >> docker/.env
-          echo -e "EXPOSE_MYSQL_PORT=${EXPOSE_MYSQL_PORT}" >> docker/.env
-          echo -e "MINIO_PORT=${MINIO_PORT}" >> docker/.env
-          echo -e "MINIO_CONSOLE_PORT=${MINIO_CONSOLE_PORT}" >> docker/.env
-          echo -e "REDIS_PORT=${REDIS_PORT}" >> docker/.env
-          echo -e "TEI_PORT=${TEI_PORT}" >> docker/.env
-          echo -e "KIBANA_PORT=${KIBANA_PORT}" >> docker/.env
-          echo -e "SVR_HTTP_PORT=${SVR_HTTP_PORT}" >> docker/.env
-          echo -e "ADMIN_SVR_HTTP_PORT=${ADMIN_SVR_HTTP_PORT}" >> docker/.env
-          echo -e "SVR_MCP_PORT=${SVR_MCP_PORT}" >> docker/.env
-          echo -e "SANDBOX_EXECUTOR_MANAGER_PORT=${SANDBOX_EXECUTOR_MANAGER_PORT}" >> docker/.env
-          echo -e "SVR_WEB_HTTP_PORT=${SVR_WEB_HTTP_PORT}" >> docker/.env
-          echo -e "SVR_WEB_HTTPS_PORT=${SVR_WEB_HTTPS_PORT}" >> docker/.env
+          # Persist computed ports into .env so docker-compose uses the correct host bindings
+               echo "" >> .env
+               echo -e "ES_PORT=${ES_PORT}" >> .env
+               echo -e "OS_PORT=${OS_PORT}" >> .env
+               echo -e "INFINITY_THRIFT_PORT=${INFINITY_THRIFT_PORT}" >> .env
+               echo -e "INFINITY_HTTP_PORT=${INFINITY_HTTP_PORT}" >> .env
+               echo -e "INFINITY_PSQL_PORT=${INFINITY_PSQL_PORT}" >> .env
+               echo -e "EXPOSE_MYSQL_PORT=${EXPOSE_MYSQL_PORT}" >> .env
+               echo -e "MINIO_PORT=${MINIO_PORT}" >> .env
+               echo -e "MINIO_CONSOLE_PORT=${MINIO_CONSOLE_PORT}" >> .env
+               echo -e "REDIS_PORT=${REDIS_PORT}" >> .env
+               echo -e "TEI_PORT=${TEI_PORT}" >> .env
+               echo -e "KIBANA_PORT=${KIBANA_PORT}" >> .env
+               echo -e "SVR_HTTP_PORT=${SVR_HTTP_PORT}" >> .env
+               echo -e "ADMIN_SVR_HTTP_PORT=${ADMIN_SVR_HTTP_PORT}" >> .env
+               echo -e "SVR_MCP_PORT=${SVR_MCP_PORT}" >> .env
+               echo -e "GO_HTTP_PORT=${GO_HTTP_PORT}" >> .env
+               echo -e "GO_ADMIN_PORT=${GO_ADMIN_PORT}" >> .env
+               echo -e "SANDBOX_EXECUTOR_MANAGER_PORT=${SANDBOX_EXECUTOR_MANAGER_PORT}" >> .env
+               echo -e "SVR_WEB_HTTP_PORT=${SVR_WEB_HTTP_PORT}" >> .env
+               echo -e "SVR_WEB_HTTPS_PORT=${SVR_WEB_HTTPS_PORT}" >> .env
+               
+               echo -e "COMPOSE_PROFILES=\${COMPOSE_PROFILES},tei-cpu" >> .env
+               echo -e "TEI_MODEL=BAAI/bge-small-en-v1.5" >> .env
+               echo -e "RAGFLOW_IMAGE=${RAGFLOW_IMAGE}" >> .env
+               echo "HOST_ADDRESS=http://host.docker.internal:${SVR_HTTP_PORT}" >> ${GITHUB_ENV}
 
-          echo -e "COMPOSE_PROFILES=\${COMPOSE_PROFILES},tei-cpu" >> docker/.env
-          echo -e "TEI_MODEL=BAAI/bge-small-en-v1.5" >> docker/.env
-          echo -e "RAGFLOW_IMAGE=${RAGFLOW_IMAGE}" >> docker/.env
-          echo "HOST_ADDRESS=http://host.docker.internal:${SVR_HTTP_PORT}" >> ${GITHUB_ENV}
+               # Patch entrypoint.sh for coverage
+               sed -i '/"\$PY" api\/ragflow_server.py \${INIT_SUPERUSER_ARGS} &/c\   echo "Ensuring coverage is installed..."\n      "$PY" -m pip install coverage -i https://mirrors.aliyun.com/pypi/simple\n     export COVERAGE_FILE=/ragflow/logs/.coverage\n        echo "Starting ragflow_server with coverage..."\n        "$PY" -m coverage run --source=./api/apps --omit="*/tests/*,*/migrations/*" -a api/ragflow_server.py ${INIT_SUPERUSER_ARGS} &' ./entrypoint.sh
+               cd ..
+               uv sync --python 3.12 --group test --frozen && uv pip install -e sdk/python
 
-          # Patch entrypoint.sh for coverage
-          sed -i '/"\$PY" api\/ragflow_server.py \${INIT_SUPERUSER_ARGS} &/c\   echo "Ensuring coverage is installed..."\n      "$PY" -m pip install coverage\n     export COVERAGE_FILE=/ragflow/logs/.coverage\n        echo "Starting ragflow_server with coverage..."\n        "$PY" -m coverage run --source=./api/apps --omit="*/tests/*,*/migrations/*" -a api/ragflow_server.py ${INIT_SUPERUSER_ARGS} &' docker/entrypoint.sh
 
+      - name: Start ragflow:nightly for Infinity
+        run: |
+          sed -i 's/^DOC_ENGINE=.*$/DOC_ENGINE=infinity/' docker/.env
+          sudo docker compose -f docker/docker-compose.yml -p ${GITHUB_RUN_ID} up -d
+
+      - name: Run sdk tests against Infinity
+        run: |
+          export http_proxy=""; export https_proxy=""; export no_proxy=""; export HTTP_PROXY=""; export HTTPS_PROXY=""; export NO_PROXY=""
+          until sudo docker exec ${RAGFLOW_CONTAINER} curl -s --connect-timeout 5 ${HOST_ADDRESS}/v1/system/ping > /dev/null 2>&1; do
+            echo "Waiting for service to be available... (last exit code: $?)"
+            sleep 5
+          done
+          echo "Start to run test sdk on Infinity"
+          source .venv/bin/activate && set -o pipefail; DOC_ENGINE=infinity pytest -s --tb=short --level=${HTTP_API_TEST_LEVEL} --junitxml=pytest-infinity-sdk.xml --cov=sdk/python/ragflow_sdk --cov-branch --cov-report=xml:coverage-infinity-sdk.xml test/testcases/test_sdk_api 2>&1 | tee infinity_sdk_test.log
+
+      - name: Run web api tests against Infinity
+        run: |
+          export http_proxy=""; export https_proxy=""; export no_proxy=""; export HTTP_PROXY=""; export HTTPS_PROXY=""; export NO_PROXY=""
+          until sudo docker exec ${RAGFLOW_CONTAINER} curl -s --connect-timeout 5 ${HOST_ADDRESS}/v1/system/ping > /dev/null 2>&1; do
+            echo "Waiting for service to be available... (last exit code: $?)"
+            sleep 5
+          done
+          source .venv/bin/activate && set -o pipefail; DOC_ENGINE=infinity pytest -s --tb=short --level=${HTTP_API_TEST_LEVEL} test/testcases/test_web_api/test_chunk_feedback 2>&1 | tee infinity_web_api_test.log
+
+      - name: Run http api tests against Infinity
+        run: |
+          export http_proxy=""; export https_proxy=""; export no_proxy=""; export HTTP_PROXY=""; export HTTPS_PROXY=""; export NO_PROXY=""
+          until sudo docker exec ${RAGFLOW_CONTAINER} curl -s --connect-timeout 5 ${HOST_ADDRESS}/v1/system/ping > /dev/null 2>&1; do
+            echo "Waiting for service to be available... (last exit code: $?)"
+            sleep 5
+          done
+          source .venv/bin/activate && set -o pipefail; DOC_ENGINE=infinity pytest -s --tb=short --level=${HTTP_API_TEST_LEVEL} test/testcases/test_http_api 2>&1 | tee infinity_http_api_test.log
+
+      - name: RAGFlow CLI retrieval test Infinity
+        env:
+          PYTHONPATH: ${{ github.workspace }}
+        run: |
+          set -euo pipefail
+          source .venv/bin/activate
+
+          export http_proxy=""; export https_proxy=""; export no_proxy=""; export HTTP_PROXY=""; export HTTPS_PROXY=""; export NO_PROXY=""
+
+          EMAIL="ci-${GITHUB_RUN_ID}@example.com"
+          PASS="ci-pass-${GITHUB_RUN_ID}"
+          DATASET="ci_dataset_${GITHUB_RUN_ID}"
+
+          CLI="python admin/client/ragflow_cli.py"
+
+          LOG_FILE="infinity_cli_test.log"
+          : > "${LOG_FILE}"
+
+          ERROR_RE='Traceback|ModuleNotFoundError|ImportError|Parse error|Bad response|Fail to|code:\\s*[1-9]'
+          run_cli() {
+            local logfile="$1"
+            shift
+            local allow_re=""
+            if [[ "${1:-}" == "--allow" ]]; then
+              allow_re="$2"
+              shift 2
+            fi
+            local cmd_display="$*"
+            echo "===== $(date -u +\"%Y-%m-%dT%H:%M:%SZ\") CMD: ${cmd_display} =====" | tee -a "${logfile}"
+            local tmp_log
+            tmp_log="$(mktemp)"
+            set +e
+            timeout 500s "$@" 2>&1 | tee "${tmp_log}"
+            local status=${PIPESTATUS[0]}
+            set -e
+            cat "${tmp_log}" >> "${logfile}"
+            if grep -qiE "${ERROR_RE}" "${tmp_log}"; then
+              if [[ -n "${allow_re}" ]] && grep -qiE "${allow_re}" "${tmp_log}"; then
+                echo "Allowed CLI error markers in ${logfile}"
+                rm -f "${tmp_log}"
+                return 0
+              fi
+              echo "Detected CLI error markers in ${logfile}"
+              rm -f "${tmp_log}"
+              exit 1
+            fi
+            rm -f "${tmp_log}"
+            return ${status}
+          }
+
+          set -a
+          source docker/.env
+          set +a
+
+          HOST_ADDRESS="http://host.docker.internal:${SVR_HTTP_PORT}"
+          USER_HOST="$(echo "${HOST_ADDRESS}" | sed -E 's#^https?://([^:/]+).*#\1#')"
+          USER_PORT="${SVR_HTTP_PORT}"
+          ADMIN_HOST="${USER_HOST}"
+          ADMIN_PORT="${ADMIN_SVR_HTTP_PORT}"
+
+          until sudo docker exec ${RAGFLOW_CONTAINER} curl -s --connect-timeout 5 ${HOST_ADDRESS}/v1/system/ping > /dev/null 2>&1; do
+            echo "Waiting for service to be available... (last exit code: $?)"
+            sleep 5
+          done
+
+          admin_ready=0
+          for i in $(seq 1 30); do
+            if run_cli "${LOG_FILE}" $CLI --type admin --host "$ADMIN_HOST" --port "$ADMIN_PORT" --username "admin@ragflow.io" --password "admin" command "ping"; then
+              admin_ready=1
+              break
+            fi
+            sleep 1
+          done
+          if [[ "${admin_ready}" -ne 1 ]]; then
+            echo "Admin service did not become ready"
+            exit 1
+          fi
+
+          run_cli "${LOG_FILE}" $CLI --type admin --host "$ADMIN_HOST" --port "$ADMIN_PORT" --username "admin@ragflow.io" --password "admin" command "show version"
+          ALLOW_USER_EXISTS_RE='already exists|already exist|duplicate|already.*registered|exist(s)?'
+          run_cli "${LOG_FILE}" --allow "${ALLOW_USER_EXISTS_RE}" $CLI --type admin --host "$ADMIN_HOST" --port "$ADMIN_PORT" --username "admin@ragflow.io" --password "admin" command "create user '$EMAIL' '$PASS'"
+
+          user_ready=0
+          for i in $(seq 1 30); do
+            if run_cli "${LOG_FILE}" $CLI --type user --host "$USER_HOST" --port "$USER_PORT" --username "$EMAIL" --password "$PASS" command "ping"; then
+              user_ready=1
+              break
+            fi
+            sleep 1
+          done
+          if [[ "${user_ready}" -ne 1 ]]; then
+            echo "User service did not become ready"
+            exit 1
+          fi
+
+          run_cli "${LOG_FILE}" $CLI --type user --host "$USER_HOST" --port "$USER_PORT" --username "$EMAIL" --password "$PASS" command "show version"
+          run_cli "${LOG_FILE}" $CLI --type user --host "$USER_HOST" --port "$USER_PORT" --username "$EMAIL" --password "$PASS" command "create dataset '$DATASET' with embedding 'BAAI/bge-small-en-v1.5@Builtin' parser 'auto'"
+          run_cli "${LOG_FILE}" $CLI --type user --host "$USER_HOST" --port "$USER_PORT" --username "$EMAIL" --password "$PASS" command "import 'test/benchmark/test_docs/Doc1.pdf,test/benchmark/test_docs/Doc2.pdf' into dataset '$DATASET'"
+          run_cli "${LOG_FILE}" $CLI --type user --host "$USER_HOST" --port "$USER_PORT" --username "$EMAIL" --password "$PASS" command "parse dataset '$DATASET' sync"
+          run_cli "${LOG_FILE}" $CLI --type user --host "$USER_HOST" --port "$USER_PORT" --username "$EMAIL" --password "$PASS" command "Benchmark 16 100 search 'what are these documents about' on datasets '$DATASET'"
+
+      - name: Stop ragflow to save coverage Infinity
+        if: ${{ !cancelled() }}
+        run: |
+          # Send SIGINT to ragflow_server.py to trigger coverage save
+          PID=$(sudo docker exec ${RAGFLOW_CONTAINER} ps aux | grep "ragflow_server.py" | grep -v grep | awk '{print $2}' | head -n 1)
+          if [ -n "$PID" ]; then
+            echo "Sending SIGINT to ragflow_server.py (PID: $PID)..."
+            sudo docker exec ${RAGFLOW_CONTAINER} kill -INT $PID
+            # Wait for process to exit and coverage file to be written
+            sleep 10
+          else
+            echo "ragflow_server.py not found!"
+          fi
+          sudo docker compose -f docker/docker-compose.yml -p ${GITHUB_RUN_ID} stop
+
+      - name: Generate server coverage report Infinity
+        if: ${{ !cancelled() }}
+        run: |
+          # .coverage file should be in docker/ragflow-logs/.coverage
+          if [ -f docker/ragflow-logs/.coverage ]; then
+            echo "Found .coverage file"
+            cp docker/ragflow-logs/.coverage .coverage
+            source .venv/bin/activate
+            # Create .coveragerc to map container paths to host paths
+            echo "[paths]" > .coveragerc
+            echo "source =" >> .coveragerc
+            echo "    ." >> .coveragerc
+            echo "    /ragflow" >> .coveragerc
+            coverage xml -o coverage-infinity-server.xml
+            rm .coveragerc
+          else
+            echo ".coverage file not found!"
+          fi
+
+      - name: Upload coverage reports to Codecov
+        uses: codecov/codecov-action@v5
+        if: ${{ !cancelled() }}
+        with:
+          token: ${{ secrets.CODECOV_TOKEN }}
+          fail_ci_if_error: false
+
+      - name: Collect ragflow log Infinity
+        if: ${{ !cancelled() }}
+        run: |
+          if [ -d docker/ragflow-logs ]; then
+            cp -r docker/ragflow-logs ${ARTIFACTS_DIR}/ragflow-logs-infinity
+            echo "ragflow log" && tail -n 200 docker/ragflow-logs/ragflow_server.log || true
+          else
+            echo "No docker/ragflow-logs directory found; skipping log collection"
+          fi
+          sudo rm -rf docker/ragflow-logs || true
+
+      - name: Stop ragflow:nightly for Infinity
+        if: always()  # always run this step even if previous steps failed
+        run: |
+          # Sometimes `docker compose down` fail due to hang container, heavy load etc. Need to remove such containers to release resources(for example, listen ports).
+          sudo docker compose -f docker/docker-compose.yml -p ${GITHUB_RUN_ID} down -v || true
+          sudo docker ps -a --filter "label=com.docker.compose.project=${GITHUB_RUN_ID}" -q | xargs -r sudo docker rm -f
+
+      - name: Start ragflow:nightly for Elasticsearch
+        run: |
+          sed -i 's/^DOC_ENGINE=.*$/DOC_ENGINE=elasticsearch/' docker/.env
           sudo docker compose -f docker/docker-compose.yml -p ${GITHUB_RUN_ID} up -d
-          uv sync --python 3.12 --group test --frozen && uv pip install -e sdk/python
 
       - name: Run sdk tests against Elasticsearch
         run: |
           export http_proxy=""; export https_proxy=""; export no_proxy=""; export HTTP_PROXY=""; export HTTPS_PROXY=""; export NO_PROXY=""
-          until sudo docker exec ${RAGFLOW_CONTAINER} curl -s --connect-timeout 5 ${HOST_ADDRESS}/v1/system/ping > /dev/null; do
-            echo "Waiting for service to be available..."
+          until sudo docker exec ${RAGFLOW_CONTAINER} curl -s --connect-timeout 5 ${HOST_ADDRESS}/v1/system/ping > /dev/null 2>&1; do
+            echo "Waiting for service to be available... (last exit code: $?)"
             sleep 5
           done
+          echo "Start to run test sdk on Elasticsearch"
           source .venv/bin/activate && set -o pipefail; pytest -s --tb=short --level=${HTTP_API_TEST_LEVEL} --junitxml=pytest-infinity-sdk.xml --cov=sdk/python/ragflow_sdk --cov-branch --cov-report=xml:coverage-es-sdk.xml test/testcases/test_sdk_api 2>&1 | tee es_sdk_test.log
 
       - name: Run web api tests against Elasticsearch
         run: |
           export http_proxy=""; export https_proxy=""; export no_proxy=""; export HTTP_PROXY=""; export HTTPS_PROXY=""; export NO_PROXY=""
-          until sudo docker exec ${RAGFLOW_CONTAINER} curl -s --connect-timeout 5 ${HOST_ADDRESS}/v1/system/ping > /dev/null; do
-            echo "Waiting for service to be available..."
+          until sudo docker exec ${RAGFLOW_CONTAINER} curl -s --connect-timeout 5 ${HOST_ADDRESS}/v1/system/ping > /dev/null 2>&1; do
+            echo "Waiting for service to be available... (last exit code: $?)"
             sleep 5
           done
           source .venv/bin/activate && set -o pipefail; pytest -s --tb=short --level=${HTTP_API_TEST_LEVEL} test/testcases/test_web_api 2>&1 | tee es_web_api_test.log
-          
+
       - name: Run http api tests against Elasticsearch
         run: |
           export http_proxy=""; export https_proxy=""; export no_proxy=""; export HTTP_PROXY=""; export HTTPS_PROXY=""; export NO_PROXY=""
-          until sudo docker exec ${RAGFLOW_CONTAINER} curl -s --connect-timeout 5 ${HOST_ADDRESS}/v1/system/ping > /dev/null; do
-            echo "Waiting for service to be available..."
+          until sudo docker exec ${RAGFLOW_CONTAINER} curl -s --connect-timeout 5 ${HOST_ADDRESS}/v1/system/ping > /dev/null 2>&1; do
+            echo "Waiting for service to be available... (last exit code: $?)"
             sleep 5
           done
           source .venv/bin/activate && set -o pipefail; pytest -s --tb=short --level=${HTTP_API_TEST_LEVEL} test/testcases/test_http_api 2>&1 | tee es_http_api_test.log
@@ -267,7 +483,7 @@ jobs:
             local tmp_log
             tmp_log="$(mktemp)"
             set +e
-            timeout 180s "$@" 2>&1 | tee "${tmp_log}"
+            timeout 500s "$@" 2>&1 | tee "${tmp_log}"
             local status=${PIPESTATUS[0]}
             set -e
             cat "${tmp_log}" >> "${logfile}"
@@ -295,8 +511,8 @@ jobs:
           ADMIN_HOST="${USER_HOST}"
           ADMIN_PORT="${ADMIN_SVR_HTTP_PORT}"
 
-          until sudo docker exec ${RAGFLOW_CONTAINER} curl -s --connect-timeout 5 ${HOST_ADDRESS}/v1/system/ping > /dev/null; do
-            echo "Waiting for service to be available..."
+          until sudo docker exec ${RAGFLOW_CONTAINER} curl -s --connect-timeout 5 ${HOST_ADDRESS}/v1/system/ping > /dev/null 2>&1; do
+            echo "Waiting for service to be available... (last exit code: $?)"
             sleep 5
           done
 
@@ -383,198 +599,7 @@ jobs:
           fi
           sudo rm -rf docker/ragflow-logs || true
 
-      - name: Stop ragflow:nightly
-        if: always()  # always run this step even if previous steps failed
-        run: |
-          sudo docker compose -f docker/docker-compose.yml -p ${GITHUB_RUN_ID} down -v || true
-          sudo docker ps -a --filter "label=com.docker.compose.project=${GITHUB_RUN_ID}" -q | xargs -r sudo docker rm -f
-
-      - name: Start ragflow:nightly
-        run: |
-          sed -i '1i DOC_ENGINE=infinity' docker/.env
-          sudo docker compose -f docker/docker-compose.yml -p ${GITHUB_RUN_ID} up -d
-
-      - name: Run sdk tests against Infinity
-        run: |
-          export http_proxy=""; export https_proxy=""; export no_proxy=""; export HTTP_PROXY=""; export HTTPS_PROXY=""; export NO_PROXY=""
-          until sudo docker exec ${RAGFLOW_CONTAINER} curl -s --connect-timeout 5 ${HOST_ADDRESS}/v1/system/ping > /dev/null; do
-            echo "Waiting for service to be available..."
-            sleep 5
-          done
-          source .venv/bin/activate && set -o pipefail; DOC_ENGINE=infinity pytest -s --tb=short --level=${HTTP_API_TEST_LEVEL} --junitxml=pytest-infinity-sdk.xml --cov=sdk/python/ragflow_sdk --cov-branch --cov-report=xml:coverage-infinity-sdk.xml test/testcases/test_sdk_api 2>&1 | tee infinity_sdk_test.log
-
-      - name: Run web api tests against Infinity
-        run: |
-          export http_proxy=""; export https_proxy=""; export no_proxy=""; export HTTP_PROXY=""; export HTTPS_PROXY=""; export NO_PROXY=""
-          until sudo docker exec ${RAGFLOW_CONTAINER} curl -s --connect-timeout 5 ${HOST_ADDRESS}/v1/system/ping > /dev/null; do
-            echo "Waiting for service to be available..."
-            sleep 5
-          done
-          source .venv/bin/activate && set -o pipefail; DOC_ENGINE=infinity pytest -s --tb=short --level=${HTTP_API_TEST_LEVEL} test/testcases/test_web_api/test_api_app 2>&1 | tee infinity_web_api_test.log
-
-      - name: Run http api tests against Infinity
-        run: |
-          export http_proxy=""; export https_proxy=""; export no_proxy=""; export HTTP_PROXY=""; export HTTPS_PROXY=""; export NO_PROXY=""
-          until sudo docker exec ${RAGFLOW_CONTAINER} curl -s --connect-timeout 5 ${HOST_ADDRESS}/v1/system/ping > /dev/null; do
-            echo "Waiting for service to be available..."
-            sleep 5
-          done
-          source .venv/bin/activate && set -o pipefail; DOC_ENGINE=infinity pytest -s --tb=short --level=${HTTP_API_TEST_LEVEL} test/testcases/test_http_api 2>&1 | tee infinity_http_api_test.log
-
-      - name: RAGFlow CLI retrieval test Infinity
-        env:
-          PYTHONPATH: ${{ github.workspace }}
-        run: |
-          set -euo pipefail
-          source .venv/bin/activate
-
-          export http_proxy=""; export https_proxy=""; export no_proxy=""; export HTTP_PROXY=""; export HTTPS_PROXY=""; export NO_PROXY=""
-
-          EMAIL="ci-${GITHUB_RUN_ID}@example.com"
-          PASS="ci-pass-${GITHUB_RUN_ID}"
-          DATASET="ci_dataset_${GITHUB_RUN_ID}"
-
-          CLI="python admin/client/ragflow_cli.py"
-
-          LOG_FILE="infinity_cli_test.log"
-          : > "${LOG_FILE}"
-
-          ERROR_RE='Traceback|ModuleNotFoundError|ImportError|Parse error|Bad response|Fail to|code:\\s*[1-9]'
-          run_cli() {
-            local logfile="$1"
-            shift
-            local allow_re=""
-            if [[ "${1:-}" == "--allow" ]]; then
-              allow_re="$2"
-              shift 2
-            fi
-            local cmd_display="$*"
-            echo "===== $(date -u +\"%Y-%m-%dT%H:%M:%SZ\") CMD: ${cmd_display} =====" | tee -a "${logfile}"
-            local tmp_log
-            tmp_log="$(mktemp)"
-            set +e
-            timeout 180s "$@" 2>&1 | tee "${tmp_log}"
-            local status=${PIPESTATUS[0]}
-            set -e
-            cat "${tmp_log}" >> "${logfile}"
-            if grep -qiE "${ERROR_RE}" "${tmp_log}"; then
-              if [[ -n "${allow_re}" ]] && grep -qiE "${allow_re}" "${tmp_log}"; then
-                echo "Allowed CLI error markers in ${logfile}"
-                rm -f "${tmp_log}"
-                return 0
-              fi
-              echo "Detected CLI error markers in ${logfile}"
-              rm -f "${tmp_log}"
-              exit 1
-            fi
-            rm -f "${tmp_log}"
-            return ${status}
-          }
-
-          set -a
-          source docker/.env
-          set +a
-
-          HOST_ADDRESS="http://host.docker.internal:${SVR_HTTP_PORT}"
-          USER_HOST="$(echo "${HOST_ADDRESS}" | sed -E 's#^https?://([^:/]+).*#\1#')"
-          USER_PORT="${SVR_HTTP_PORT}"
-          ADMIN_HOST="${USER_HOST}"
-          ADMIN_PORT="${ADMIN_SVR_HTTP_PORT}"
-
-          until sudo docker exec ${RAGFLOW_CONTAINER} curl -s --connect-timeout 5 ${HOST_ADDRESS}/v1/system/ping > /dev/null; do
-            echo "Waiting for service to be available..."
-            sleep 5
-          done
-
-          admin_ready=0
-          for i in $(seq 1 30); do
-            if run_cli "${LOG_FILE}" $CLI --type admin --host "$ADMIN_HOST" --port "$ADMIN_PORT" --username "admin@ragflow.io" --password "admin" command "ping"; then
-              admin_ready=1
-              break
-            fi
-            sleep 1
-          done
-          if [[ "${admin_ready}" -ne 1 ]]; then
-            echo "Admin service did not become ready"
-            exit 1
-          fi
-
-          run_cli "${LOG_FILE}" $CLI --type admin --host "$ADMIN_HOST" --port "$ADMIN_PORT" --username "admin@ragflow.io" --password "admin" command "show version"
-          ALLOW_USER_EXISTS_RE='already exists|already exist|duplicate|already.*registered|exist(s)?'
-          run_cli "${LOG_FILE}" --allow "${ALLOW_USER_EXISTS_RE}" $CLI --type admin --host "$ADMIN_HOST" --port "$ADMIN_PORT" --username "admin@ragflow.io" --password "admin" command "create user '$EMAIL' '$PASS'"
-
-          user_ready=0
-          for i in $(seq 1 30); do
-            if run_cli "${LOG_FILE}" $CLI --type user --host "$USER_HOST" --port "$USER_PORT" --username "$EMAIL" --password "$PASS" command "ping"; then
-              user_ready=1
-              break
-            fi
-            sleep 1
-          done
-          if [[ "${user_ready}" -ne 1 ]]; then
-            echo "User service did not become ready"
-            exit 1
-          fi
-
-          run_cli "${LOG_FILE}" $CLI --type user --host "$USER_HOST" --port "$USER_PORT" --username "$EMAIL" --password "$PASS" command "show version"
-          run_cli "${LOG_FILE}" $CLI --type user --host "$USER_HOST" --port "$USER_PORT" --username "$EMAIL" --password "$PASS" command "create dataset '$DATASET' with embedding 'BAAI/bge-small-en-v1.5@Builtin' parser 'auto'"
-          run_cli "${LOG_FILE}" $CLI --type user --host "$USER_HOST" --port "$USER_PORT" --username "$EMAIL" --password "$PASS" command "import 'test/benchmark/test_docs/Doc1.pdf,test/benchmark/test_docs/Doc2.pdf' into dataset '$DATASET'"
-          run_cli "${LOG_FILE}" $CLI --type user --host "$USER_HOST" --port "$USER_PORT" --username "$EMAIL" --password "$PASS" command "parse dataset '$DATASET' sync"
-          run_cli "${LOG_FILE}" $CLI --type user --host "$USER_HOST" --port "$USER_PORT" --username "$EMAIL" --password "$PASS" command "Benchmark 16 100 search 'what are these documents about' on datasets '$DATASET'"
-
-      - name: Stop ragflow to save coverage Infinity
-        if: ${{ !cancelled() }}
-        run: |
-          # Send SIGINT to ragflow_server.py to trigger coverage save
-          PID=$(sudo docker exec ${RAGFLOW_CONTAINER} ps aux | grep "ragflow_server.py" | grep -v grep | awk '{print $2}' | head -n 1)
-          if [ -n "$PID" ]; then
-            echo "Sending SIGINT to ragflow_server.py (PID: $PID)..."
-            sudo docker exec ${RAGFLOW_CONTAINER} kill -INT $PID
-            # Wait for process to exit and coverage file to be written
-            sleep 10
-          else
-            echo "ragflow_server.py not found!"
-          fi
-          sudo docker compose -f docker/docker-compose.yml -p ${GITHUB_RUN_ID} stop
-
-      - name: Generate server coverage report Infinity
-        if: ${{ !cancelled() }}
-        run: |
-          # .coverage file should be in docker/ragflow-logs/.coverage
-          if [ -f docker/ragflow-logs/.coverage ]; then
-            echo "Found .coverage file"
-            cp docker/ragflow-logs/.coverage .coverage
-            source .venv/bin/activate
-            # Create .coveragerc to map container paths to host paths
-            echo "[paths]" > .coveragerc
-            echo "source =" >> .coveragerc
-            echo "    ." >> .coveragerc
-            echo "    /ragflow" >> .coveragerc
-            coverage xml -o coverage-infinity-server.xml
-            rm .coveragerc
-          else
-            echo ".coverage file not found!"
-          fi
-          
-      - name: Upload coverage reports to Codecov
-        uses: codecov/codecov-action@v5
-        if: ${{ !cancelled() }}
-        with:
-          token: ${{ secrets.CODECOV_TOKEN }}
-          fail_ci_if_error: false
-
-      - name: Collect ragflow log
-        if: ${{ !cancelled() }}
-        run: |
-          if [ -d docker/ragflow-logs ]; then
-            cp -r docker/ragflow-logs ${ARTIFACTS_DIR}/ragflow-logs-infinity
-            echo "ragflow log" && tail -n 200 docker/ragflow-logs/ragflow_server.log || true
-          else
-            echo "No docker/ragflow-logs directory found; skipping log collection"
-          fi
-          sudo rm -rf docker/ragflow-logs || true
-
-      - name: Stop ragflow:nightly
+      - name: Stop ragflow:nightly for Elasticsearch
         if: always()  # always run this step even if previous steps failed
         run: |
           # Sometimes `docker compose down` fail due to hang container, heavy load etc. Need to remove such containers to release resources(for example, listen ports).

.gitignore

@@ -7,7 +7,7 @@ hudet/
 cv/
 layout_app.py
 api/flask_session
-
+venv/
 # Remove Cargo.lock from gitignore if creating an executable, leave it for libraries
 # More information here https://doc.rust-lang.org/cargo/guide/cargo-toml-vs-cargo-lock.html
 Cargo.lock
@@ -205,9 +205,29 @@ ragflow_cli.egg-info
 backup
 
 
+*huqie.txt
+
 .hypothesis
 
 
 # Added by cargo
 
 /target
+
+# Do not include in PR (local dev / build artifacts)
+ragflow.egg-info/
+uv-aarch64*.tar.gz
+uv-aarch64-unknown-linux-gnu.tar.gz
+docker/launch_backend_service_windows.sh
+
+# C++ build directories
+internal/cpp/build/
+internal/cpp/cmake-build-release/
+internal/cpp/cmake-build-debug/
+
+# Trae IDE config
+.trae/
+
+# Go server build output
+bin/*
+!bin/.gitkeep

CLAUDE.md

@@ -5,14 +5,16 @@ This file provides guidance to Claude Code (claude.ai/code) when working with co
 ## Project Overview
 
 RAGFlow is an open-source RAG (Retrieval-Augmented Generation) engine based on deep document understanding. It's a full-stack application with:
+
 - Python backend (Flask-based API server)
-- React/TypeScript frontend (built with UmiJS)
+- React/TypeScript frontend (built with vitejs)
 - Microservices architecture with Docker deployment
 - Multiple data stores (MySQL, Elasticsearch/Infinity, Redis, MinIO)
 
 ## Architecture
 
 ### Backend (`/api/`)
+
 - **Main Server**: `api/ragflow_server.py` - Flask application entry point
 - **Apps**: Modular Flask blueprints in `api/apps/` for different functionalities:
   - `kb_app.py` - Knowledge base management
@@ -24,25 +26,29 @@ RAGFlow is an open-source RAG (Retrieval-Augmented Generation) engine based on d
 - **Models**: Database models in `api/db/db_models.py`
 
 ### Core Processing (`/rag/`)
+
 - **Document Processing**: `deepdoc/` - PDF parsing, OCR, layout analysis
 - **LLM Integration**: `rag/llm/` - Model abstractions for chat, embedding, reranking
 - **RAG Pipeline**: `rag/flow/` - Chunking, parsing, tokenization
 - **Graph RAG**: `rag/graphrag/` - Knowledge graph construction and querying
 
 ### Agent System (`/agent/`)
+
 - **Components**: Modular workflow components (LLM, retrieval, categorize, etc.)
 - **Templates**: Pre-built agent workflows in `agent/templates/`
 - **Tools**: External API integrations (Tavily, Wikipedia, SQL execution, etc.)
 
 ### Frontend (`/web/`)
-- React/TypeScript with UmiJS framework
-- Ant Design + shadcn/ui components
+
+- React/TypeScript with vitejs framework
+- shadcn/ui components
 - State management with Zustand
 - Tailwind CSS for styling
 
 ## Common Development Commands
 
 ### Backend Development
+
 ```bash
 # Install Python dependencies
 uv sync --python 3.12 --all-extras
@@ -66,6 +72,7 @@ ruff format
 ```
 
 ### Frontend Development
+
 ```bash
 cd web
 npm install
@@ -76,6 +83,7 @@ npm run test       # Jest tests
 ```
 
 ### Docker Development
+
 ```bash
 # Full stack with Docker
 cd docker
@@ -104,6 +112,7 @@ docker build --platform linux/amd64 -f Dockerfile -t infiniflow/ragflow:nightly
 ## Database Engines
 
 RAGFlow supports switching between Elasticsearch (default) and Infinity:
+
 - Set `DOC_ENGINE=infinity` in `docker/.env` to use Infinity
 - Requires container restart: `docker compose down -v && docker compose up -d`
 
@@ -114,3 +123,12 @@ RAGFlow supports switching between Elasticsearch (default) and Infinity:
 - Docker & Docker Compose
 - uv package manager
 - 16GB+ RAM, 50GB+ disk space
+
+1. Think before acting. Read existing files before writing code.
+2. Be concise in output but thorough in reasoning.
+3. Prefer editing over rewriting whole files.
+4. Do not re-read files you have already read.
+5. Test your code before declaring done.
+6. No sycophantic openers or closing fluff.
+7. Keep solutions simple and direct.
+8. User instructions always override this file.

Dockerfile

@@ -7,7 +7,7 @@ ARG NEED_MIRROR=0
 
 WORKDIR /ragflow
 
-# Copy models downloaded via download_deps.py
+# copy models downloaded via download_deps.py
 RUN mkdir -p /ragflow/rag/res/deepdoc /root/.ragflow
 RUN --mount=type=bind,from=infiniflow/ragflow_deps:latest,source=/huggingface.co,target=/huggingface.co \
     tar --exclude='.*' -cf - \
@@ -19,49 +19,49 @@ RUN --mount=type=bind,from=infiniflow/ragflow_deps:latest,source=/huggingface.co
 # This is the only way to run python-tika without internet access. Without this set, the default is to check the tika version and pull latest every time from Apache.
 RUN --mount=type=bind,from=infiniflow/ragflow_deps:latest,source=/,target=/deps \
     cp -r /deps/nltk_data /root/ && \
-    cp /deps/tika-server-standard-3.2.3.jar /deps/tika-server-standard-3.2.3.jar.md5 /ragflow/ && \
+    cp /deps/tika-server-standard-3.3.0.jar /deps/tika-server-standard-3.3.0.jar.md5 /ragflow/ && \
     cp /deps/cl100k_base.tiktoken /ragflow/9b5ad71b2ce5302211f9c61530b329a4922fc6a4
 
-ENV TIKA_SERVER_JAR="file:///ragflow/tika-server-standard-3.2.3.jar"
+ENV TIKA_SERVER_JAR="file:///ragflow/tika-server-standard-3.3.0.jar"
 ENV DEBIAN_FRONTEND=noninteractive
 
 # Setup apt
 # Python package and implicit dependencies:
 # opencv-python: libglib2.0-0 libglx-mesa0 libgl1
-# python-pptx:   default-jdk                              tika-server-standard-3.2.3.jar
+# python-pptx:   default-jdk                              tika-server-standard-3.3.0.jar
 # selenium:      libatk-bridge2.0-0                       chrome-linux64-121-0-6167-85
 # Building C extensions: libpython3-dev libgtk-4-1 libnss3 xdg-utils libgbm-dev
 RUN --mount=type=cache,id=ragflow_apt,target=/var/cache/apt,sharing=locked \
     apt update && \
     apt --no-install-recommends install -y ca-certificates; \
     if [ "$NEED_MIRROR" == "1" ]; then \
-        sed -i 's|http://archive.ubuntu.com/ubuntu|https://mirrors.tuna.tsinghua.edu.cn/ubuntu|g' /etc/apt/sources.list.d/ubuntu.sources; \
-        sed -i 's|http://security.ubuntu.com/ubuntu|https://mirrors.tuna.tsinghua.edu.cn/ubuntu|g' /etc/apt/sources.list.d/ubuntu.sources; \
+        sed -i 's|http://archive.ubuntu.com/ubuntu|https://mirrors.aliyun.com/ubuntu|g' /etc/apt/sources.list.d/ubuntu.sources; \
+        sed -i 's|http://security.ubuntu.com/ubuntu|https://mirrors.aliyun.com/ubuntu|g' /etc/apt/sources.list.d/ubuntu.sources; \
     fi; \
     rm -f /etc/apt/apt.conf.d/docker-clean && \
     echo 'Binary::apt::APT::Keep-Downloaded-Packages "true";' > /etc/apt/apt.conf.d/keep-cache && \
     chmod 1777 /tmp && \
     apt update && \
-    apt install -y libglib2.0-0 libglx-mesa0 libgl1 && \
-    apt install -y pkg-config libicu-dev libgdiplus && \
-    apt install -y default-jdk && \
-    apt install -y libatk-bridge2.0-0 && \
-    apt install -y libpython3-dev libgtk-4-1 libnss3 xdg-utils libgbm-dev && \
-    apt install -y libjemalloc-dev && \
-    apt install -y gnupg unzip curl wget git vim less && \
-    apt install -y ghostscript && \
-    apt install -y pandoc && \
-    apt install -y texlive && \
-    apt install -y fonts-freefont-ttf fonts-noto-cjk && \
-    apt install -y postgresql-client
+    apt install -y \
+    build-essential libglib2.0-0 libglx-mesa0 libgl1 pkg-config libicu-dev libgdiplus default-jdk libatk-bridge2.0-0 libpython3-dev libgtk-4-1 libnss3 xdg-utils libgbm-dev libjemalloc-dev gnupg unzip curl wget git vim less ghostscript pandoc texlive texlive-latex-extra texlive-xetex texlive-lang-chinese fonts-freefont-ttf fonts-noto-cjk postgresql-client
+
+# Download resource from GitHub to /usr/share/infinity
+RUN mkdir -p /usr/share/infinity/resource && \
+    if [ "$NEED_MIRROR" == "1" ]; then \
+        git clone --depth 1 --single-branch https://gitee.com/infiniflow/resource /tmp/resource; \
+    else \
+        git clone --depth 1 --single-branch https://github.com/infiniflow/resource.git /tmp/resource; \
+    fi && \
+    cp -r /tmp/resource/* /usr/share/infinity/resource && \
+    rm -rf /tmp/resource
 
 ARG NGINX_VERSION=1.29.5-1~noble
 RUN --mount=type=cache,id=ragflow_apt,target=/var/cache/apt,sharing=locked \
     mkdir -p /etc/apt/keyrings && \
-    curl -fsSL https://nginx.org/keys/nginx_signing.key | gpg --dearmor -o /etc/apt/keyrings/nginx-archive-keyring.gpg && \
+    curl --retry 5 --retry-delay 2 --retry-all-errors -fsSL https://nginx.org/keys/nginx_signing.key | gpg --dearmor -o /etc/apt/keyrings/nginx-archive-keyring.gpg && \
     echo "deb [signed-by=/etc/apt/keyrings/nginx-archive-keyring.gpg] https://nginx.org/packages/mainline/ubuntu/ noble nginx" > /etc/apt/sources.list.d/nginx.list && \
-    apt update && \
-    apt install -y nginx=${NGINX_VERSION} && \
+    apt -o Acquire::Retries=5 update && \
+    apt -o Acquire::Retries=5 install -y nginx=${NGINX_VERSION} && \
     apt-mark hold nginx
 
 # Install uv
@@ -70,7 +70,7 @@ RUN --mount=type=bind,from=infiniflow/ragflow_deps:latest,source=/,target=/deps
         mkdir -p /etc/uv && \
         echo 'python-install-mirror = "https://registry.npmmirror.com/-/binary/python-build-standalone/"' > /etc/uv/uv.toml && \
         echo '[[index]]' >> /etc/uv/uv.toml && \
-        echo 'url = "https://pypi.tuna.tsinghua.edu.cn/simple"' >> /etc/uv/uv.toml && \
+        echo 'url = "https://mirrors.aliyun.com/pypi/simple"' >> /etc/uv/uv.toml && \
         echo 'default = true' >> /etc/uv/uv.toml; \
     fi; \
     arch="$(uname -m)"; \
@@ -80,33 +80,19 @@ RUN --mount=type=bind,from=infiniflow/ragflow_deps:latest,source=/,target=/deps
     && rm -rf "uv-${uv_arch}-unknown-linux-gnu" \
     && uv python install 3.12
 
-ENV PYTHONDONTWRITEBYTECODE=1 DOTNET_SYSTEM_GLOBALIZATION_INVARIANT=1
+ENV PYTHONDONTWRITEBYTECODE=1 DOTNET_SYSTEM_GLOBALIZATION_INVARIANT=1 \
+    UV_HTTP_TIMEOUT=200 \
+    UV_HTTP_RETRIES=3
 ENV PATH=/root/.local/bin:$PATH
 
 # nodejs 12.22 on Ubuntu 22.04 is too old
 RUN --mount=type=cache,id=ragflow_apt,target=/var/cache/apt,sharing=locked \
     curl -fsSL https://deb.nodesource.com/setup_20.x | bash - && \
-    apt purge -y nodejs npm cargo && \
+    apt purge -y nodejs npm && \
     apt autoremove -y && \
     apt update && \
     apt install -y nodejs
 
-# A modern version of cargo is needed for the latest version of the Rust compiler.
-RUN apt update && apt install -y curl build-essential \
-    && if [ "$NEED_MIRROR" == "1" ]; then \
-         # Use TUNA mirrors for rustup/rust dist files \
-         export RUSTUP_DIST_SERVER="https://mirrors.tuna.tsinghua.edu.cn/rustup"; \
-         export RUSTUP_UPDATE_ROOT="https://mirrors.tuna.tsinghua.edu.cn/rustup/rustup"; \
-         echo "Using TUNA mirrors for Rustup."; \
-       fi; \
-    # Force curl to use HTTP/1.1 \
-    curl --proto '=https' --tlsv1.2 --http1.1 -sSf https://sh.rustup.rs | bash -s -- -y --profile minimal \
-    && echo 'export PATH="/root/.cargo/bin:${PATH}"' >> /root/.bashrc
-
-ENV PATH="/root/.cargo/bin:${PATH}"
-
-RUN cargo --version && rustc --version
-
 # Add msssql ODBC driver
 # macOS ARM64 environment, install msodbcsql18.
 # general x86_64 environment, install msodbcsql17.
@@ -157,9 +143,9 @@ COPY pyproject.toml uv.lock ./
 # uv records index url into uv.lock but doesn't failover among multiple indexes
 RUN --mount=type=cache,id=ragflow_uv,target=/root/.cache/uv,sharing=locked \
     if [ "$NEED_MIRROR" == "1" ]; then \
-        sed -i 's|pypi.org|pypi.tuna.tsinghua.edu.cn|g' uv.lock; \
+        sed -i 's|pypi.org|mirrors.aliyun.com/pypi|g' uv.lock; \
     else \
-        sed -i 's|pypi.tuna.tsinghua.edu.cn|pypi.org|g' uv.lock; \
+        sed -i 's|mirrors.aliyun.com/pypi|pypi.org|g' uv.lock; \
     fi; \
     uv sync --python 3.12 --frozen && \
     # Ensure pip is available in the venv for runtime package installation (fixes #12651)
@@ -168,8 +154,8 @@ RUN --mount=type=cache,id=ragflow_uv,target=/root/.cache/uv,sharing=locked \
 COPY web web
 COPY docs docs
 RUN --mount=type=cache,id=ragflow_npm,target=/root/.npm,sharing=locked \
-    export NODE_OPTIONS="--max-old-space-size=4096" && \
-    cd web && npm install && npm run build
+    cd web && NODE_OPTIONS="--max-old-space-size=8192" npm install && \
+    NODE_OPTIONS="--max-old-space-size=8192" VITE_BUILD_SOURCEMAP=false VITE_MINIFY=esbuild npm run build
 
 COPY .git /ragflow/.git
 
@@ -202,11 +188,19 @@ COPY pyproject.toml uv.lock ./
 COPY mcp mcp
 COPY common common
 COPY memory memory
+COPY bin bin
 
 COPY docker/service_conf.yaml.template ./conf/service_conf.yaml.template
 COPY docker/entrypoint.sh ./
 RUN chmod +x ./entrypoint*.sh
 
+# Copy nginx configuration for frontend serving
+COPY docker/nginx/ragflow.conf.golang docker/nginx/ragflow.conf.python docker/nginx/ragflow.conf.hybrid docker/nginx/nginx.conf docker/nginx/proxy.conf /etc/nginx/
+RUN mv /etc/nginx/ragflow.conf.golang /etc/nginx/conf.d/ragflow.conf.golang && \
+    mv /etc/nginx/ragflow.conf.python /etc/nginx/conf.d/ragflow.conf.python && \
+    mv /etc/nginx/ragflow.conf.hybrid /etc/nginx/conf.d/ragflow.conf.hybrid && \
+    rm -f /etc/nginx/sites-enabled/default
+
 # Copy compiled web pages
 COPY --from=builder /ragflow/web/dist /ragflow/web/dist
 

Dockerfile.deps

@@ -3,7 +3,7 @@
 FROM scratch
 
 # Copy resources downloaded via download_deps.py
-COPY chromedriver-linux64-121-0-6167-85 chrome-linux64-121-0-6167-85 cl100k_base.tiktoken libssl1.1_1.1.1f-1ubuntu2_amd64.deb libssl1.1_1.1.1f-1ubuntu2_arm64.deb tika-server-standard-3.2.3.jar tika-server-standard-3.2.3.jar.md5 libssl*.deb uv-x86_64-unknown-linux-gnu.tar.gz uv-aarch64-unknown-linux-gnu.tar.gz /
+COPY chromedriver-linux64-121-0-6167-85 chrome-linux64-121-0-6167-85 cl100k_base.tiktoken libssl1.1_1.1.1f-1ubuntu2_amd64.deb libssl1.1_1.1.1f-1ubuntu2_arm64.deb tika-server-standard-3.3.0.jar tika-server-standard-3.3.0.jar.md5 libssl*.deb uv-x86_64-unknown-linux-gnu.tar.gz uv-aarch64-unknown-linux-gnu.tar.gz /
 
 COPY nltk_data /nltk_data
 

README.md

@@ -1,5 +1,5 @@
 <div align="center">
-<a href="https://demo.ragflow.io/">
+<a href="https://cloud.ragflow.io/">
 <img src="web/src/assets/logo-with-text.svg" width="520" alt="ragflow logo">
 </a>
 </div>
@@ -12,17 +12,20 @@
   <a href="./README_ko.md"><img alt="한국어" src="https://img.shields.io/badge/한국어-DFE0E5"></a>
   <a href="./README_id.md"><img alt="Bahasa Indonesia" src="https://img.shields.io/badge/Bahasa Indonesia-DFE0E5"></a>
   <a href="./README_pt_br.md"><img alt="Português(Brasil)" src="https://img.shields.io/badge/Português(Brasil)-DFE0E5"></a>
+  <a href="./README_fr.md"><img alt="README en Français" src="https://img.shields.io/badge/Français-DFE0E5"></a>
+  <a href="./README_ar.md"><img alt="README in Arabic" src="https://img.shields.io/badge/Arabic-DFE0E5"></a>
+  <a href="./README_tr.md"><img alt="Türkçe README" src="https://img.shields.io/badge/Türkçe-DFE0E5"></a>
 </p>
 
 <p align="center">
     <a href="https://x.com/intent/follow?screen_name=infiniflowai" target="_blank">
         <img src="https://img.shields.io/twitter/follow/infiniflow?logo=X&color=%20%23f5f5f5" alt="follow on X(Twitter)">
     </a>
-    <a href="https://demo.ragflow.io" target="_blank">
+    <a href="https://cloud.ragflow.io" target="_blank">
         <img alt="Static Badge" src="https://img.shields.io/badge/Online-Demo-4e6b99">
     </a>
     <a href="https://hub.docker.com/r/infiniflow/ragflow" target="_blank">
-        <img src="https://img.shields.io/docker/pulls/infiniflow/ragflow?label=Docker%20Pulls&color=0db7ed&logo=docker&logoColor=white&style=flat-square" alt="docker pull infiniflow/ragflow:v0.24.0">
+        <img src="https://img.shields.io/docker/pulls/infiniflow/ragflow?label=Docker%20Pulls&color=0db7ed&logo=docker&logoColor=white&style=flat-square" alt="docker pull infiniflow/ragflow:v0.25.0">
     </a>
     <a href="https://github.com/infiniflow/ragflow/releases/latest">
         <img src="https://img.shields.io/github/v/release/infiniflow/ragflow?color=blue&label=Latest%20Release" alt="Latest Release">
@@ -40,7 +43,7 @@
   <a href="https://github.com/infiniflow/ragflow/issues/12241">Roadmap</a> |
   <a href="https://twitter.com/infiniflowai">Twitter</a> |
   <a href="https://discord.gg/NjYzJD3GM3">Discord</a> |
-  <a href="https://demo.ragflow.io">Demo</a>
+  <a href="https://cloud.ragflow.io">Demo</a>
 </h4>
 
 <div align="center" style="margin-top:20px;margin-bottom:20px;">
@@ -76,7 +79,7 @@
 
 ## 🎮 Demo
 
-Try our demo at [https://demo.ragflow.io](https://demo.ragflow.io).
+Try our demo at [https://cloud.ragflow.io](https://cloud.ragflow.io).
 
 <div align="center" style="margin-top:20px;margin-bottom:20px;">
 <img src="https://raw.githubusercontent.com/infiniflow/ragflow-docs/refs/heads/image/image/chunking.gif" width="1200"/>
@@ -85,6 +88,7 @@ Try our demo at [https://demo.ragflow.io](https://demo.ragflow.io).
 
 ## 🔥 Latest Updates
 
+- 2026-03-24 [RAGFlow Skill on OpenClaw](https://clawhub.ai/yingfeng/ragflow-skill) — Provides an official skill for accessing RAGFlow datasets via OpenClaw.
 - 2025-12-26 Supports 'Memory' for AI agent.
 - 2025-11-19 Supports Gemini 3 Pro.
 - 2025-11-12 Supports data synchronization from Confluence, S3, Notion, Discord, Google Drive.
@@ -188,12 +192,12 @@ releases! 🌟
 > All Docker images are built for x86 platforms. We don't currently offer Docker images for ARM64.
 > If you are on an ARM64 platform, follow [this guide](https://ragflow.io/docs/dev/build_docker_image) to build a Docker image compatible with your system.
 
-> The command below downloads the `v0.24.0` edition of the RAGFlow Docker image. See the following table for descriptions of different RAGFlow editions. To download a RAGFlow edition different from `v0.24.0`, update the `RAGFLOW_IMAGE` variable accordingly in **docker/.env** before using `docker compose` to start the server.
+> The command below downloads the `v0.25.0` edition of the RAGFlow Docker image. See the following table for descriptions of different RAGFlow editions. To download a RAGFlow edition different from `v0.25.0`, update the `RAGFLOW_IMAGE` variable accordingly in **docker/.env** before using `docker compose` to start the server.
 
 ```bash
    $ cd ragflow/docker
 
-   # git checkout v0.24.0
+   # git checkout v0.25.0
    # Optional: use a stable tag (see releases: https://github.com/infiniflow/ragflow/releases)
    # This step ensures the **entrypoint.sh** file in the code matches the Docker image version.
 
@@ -325,7 +329,7 @@ docker build --platform linux/amd64 \
    git clone https://github.com/infiniflow/ragflow.git
    cd ragflow/
    uv sync --python 3.12 # install RAGFlow dependent python modules
-   uv run download_deps.py
+   uv run python3 download_deps.py
    pre-commit install
    ```
 3. Launch the dependent services (MinIO, Elasticsearch, Redis, and MySQL) using Docker Compose:
@@ -389,8 +393,8 @@ docker build --platform linux/amd64 \
 - [Quickstart](https://ragflow.io/docs/dev/)
 - [Configuration](https://ragflow.io/docs/dev/configurations)
 - [Release notes](https://ragflow.io/docs/dev/release_notes)
-- [User guides](https://ragflow.io/docs/dev/category/guides)
-- [Developer guides](https://ragflow.io/docs/dev/category/developers)
+- [User guides](https://ragflow.io/docs/category/user-guides)
+- [Developer guides](https://ragflow.io/docs/category/developer-guides)
 - [References](https://ragflow.io/docs/dev/category/references)
 - [FAQs](https://ragflow.io/docs/dev/faq)
 

README_ar.md

@@ -0,0 +1,414 @@
+<div align="center">
+<a href="https://cloud.ragflow.io/">
+<img src="web/src/assets/logo-with-text.svg" width="520" alt="ragflow logo">
+</a>
+</div>
+
+<p align="center">
+  <a href="./README.md"><img alt="README in English" src="https://img.shields.io/badge/English-DFE0E5"></a>
+  <a href="./README_zh.md"><img alt="简体中文版自述文件" src="https://img.shields.io/badge/简体中文-DFE0E5"></a>
+  <a href="./README_tzh.md"><img alt="繁體版中文自述文件" src="https://img.shields.io/badge/繁體中文-DFE0E5"></a>
+  <a href="./README_ja.md"><img alt="日本語のREADME" src="https://img.shields.io/badge/日本語-DFE0E5"></a>
+  <a href="./README_ko.md"><img alt="한국어" src="https://img.shields.io/badge/한국어-DFE0E5"></a>
+  <a href="./README_id.md"><img alt="Bahasa Indonesia" src="https://img.shields.io/badge/Bahasa Indonesia-DFE0E5"></a>
+  <a href="./README_pt_br.md"><img alt="Português(Brasil)" src="https://img.shields.io/badge/Português(Brasil)-DFE0E5"></a>
+  <a href="./README_fr.md"><img alt="README en Français" src="https://img.shields.io/badge/Français-DFE0E5"></a>
+  <a href="./README_ar.md"><img alt="README in Arabic" src="https://img.shields.io/badge/Arabic-DBEDFA"></a>
+  <a href="./README_tr.md"><img alt="Türkçe README" src="https://img.shields.io/badge/Türkçe-DFE0E5"></a>
+</p>
+
+<p align="center">
+    <a href="https://x.com/intent/follow?screen_name=infiniflowai" target="_blank">
+        <img src="https://img.shields.io/twitter/follow/infiniflow?logo=X&color=%20%23f5f5f5" alt="follow on X(Twitter)">
+    </a>
+    <a href="https://cloud.ragflow.io" target="_blank">
+        <img alt="Static Badge" src="https://img.shields.io/badge/Online-Demo-4e6b99">
+    </a>
+    <a href="https://hub.docker.com/r/infiniflow/ragflow" target="_blank">
+        <img src="https://img.shields.io/docker/pulls/infiniflow/ragflow?label=Docker%20Pulls&color=0db7ed&logo=docker&logoColor=white&style=flat-square" alt="docker pull infiniflow/ragflow:v0.25.0">
+    </a>
+    <a href="https://github.com/infiniflow/ragflow/releases/latest">
+        <img src="https://img.shields.io/github/v/release/infiniflow/ragflow?color=blue&label=Latest%20Release" alt="Latest Release">
+    </a>
+    <a href="https://github.com/infiniflow/ragflow/blob/main/LICENSE">
+        <img height="21" src="https://img.shields.io/badge/License-Apache--2.0-ffffff?labelColor=d4eaf7&color=2e6cc4" alt="license">
+    </a>
+    <a href="https://deepwiki.com/infiniflow/ragflow">
+        <img alt="Ask DeepWiki" src="https://deepwiki.com/badge.svg">
+    </a>
+</p>
+
+<h4 align="center">
+  <a href="https://ragflow.io/docs/dev/">Document</a> |
+  <a href="https://github.com/infiniflow/ragflow/issues/12241">Roadmap</a> |
+  <a href="https://twitter.com/infiniflowai">Twitter</a> |
+  <a href="https://discord.gg/NjYzJD3GM3">Discord</a> |
+  <a href="https://cloud.ragflow.io">Demo</a>
+</h4>
+
+<div align="center" style="margin-top:20px;margin-bottom:20px;">
+<img src="https://raw.githubusercontent.com/infiniflow/ragflow-docs/refs/heads/image/image/ragflow-octoverse.png" width="1200"/>
+</div>
+
+<div align="center">
+<a href="https://trendshift.io/repositories/9064" target="_blank"><img src="https://trendshift.io/api/badge/repositories/9064" alt="infiniflow%2Fragflow | Trendshift" style="width: 250px; height: 55px;" width="250" height="55"/></a>
+</div>
+
+<details open>
+<summary><b>📕 جدول المحتويات</b></summary>
+
+- 💡 [ما هو RAGFlow؟](#-what-is-ragflow)
+- 🎮 [Demo](#-demo)
+- 📌 [آخر التحديثات](#-latest-updates)
+- 🌟 [الميزات الرئيسية](#-key-features)
+- 🔎 [بنية النظام](#-system-architecture)
+- 🎬 [ابدأ](#-get-started)
+- 🔧 [التكوينات](#-configurations)
+- 🔧 [إنشاء صورة Docker](#-build-a-docker-image)
+- 🔨 [إطلاق الخدمة من المصدر للتطوير](#-launch-service-from-source-for-development)
+- 📚 [التوثيق](#-documentation)
+- 📜 [Roadmap](#-roadmap)
+- 🏄 [المجتمع](#-community)
+- 🙌 [مساهمة](#-contributing)
+
+</details>
+
+## 💡 ما هو RAGFlow؟
+
+يُعد مشروع [RAGFlow](https://ragflow.io/) محركًا رائدًا ومفتوح المصدر للاسترجاع المعزز بالتوليد (<bdi dir="ltr">RAG</bdi>)، ويجمع أحدث تقنيات <bdi dir="ltr">RAG</bdi> مع قدرات الوكلاء لبناء طبقة سياق متقدمة لنماذج <bdi dir="ltr">LLMs</bdi>. يوفّر سير عمل <bdi dir="ltr">RAG</bdi> مبسّطًا وقابلًا للتكيّف مع المؤسسات بمختلف أحجامها. وبالاعتماد على [محرك سياق موحّد](https://ragflow.io/basics/what-is-agent-context-engine) وقوالب وكلاء جاهزة، يتيح <bdi dir="ltr">RAGFlow</bdi> للمطورين تحويل البيانات المعقّدة إلى أنظمة <bdi dir="ltr">AI</bdi> عالية الدقة وجاهزة للإنتاج بكفاءة وموثوقية.
+
+## 🎮 Demo
+
+جرّب النسخة التجريبية على [https://cloud.ragflow.io](https://cloud.ragflow.io).
+
+<div align="center" style="margin-top:20px;margin-bottom:20px;">
+<img src="https://raw.githubusercontent.com/infiniflow/ragflow-docs/refs/heads/image/image/chunking.gif" width="1200"/>
+<img src="https://raw.githubusercontent.com/infiniflow/ragflow-docs/refs/heads/image/image/agentic-dark.gif" width="1200"/>
+</div>
+
+## 🔥 آخر التحديثات
+
+- 2026-03-24 [RAGFlow Skill on OpenClaw](https://clawhub.ai/yingfeng/ragflow-skill) — توفر مهارة رسمية للوصول إلى مجموعات بيانات RAGFlow عبر OpenClaw.
+- 2025-12-26 يدعم ميزة "Memory" لوكلاء الذكاء الاصطناعي.
+- 11-11-2025 يدعم Gemini 3 Pro.
+- 12-11-2025 يدعم مزامنة البيانات من Confluence، S3، Notion، Discord، Google Drive.
+- 23-10-2025 يدعم MinerU وDocling كطرق لتحليل المستندات.
+- 15-10-2025 يدعم العرض الأوركسترالي pipeline.
+- 08-08-2025 يدعم أحدث موديلات سلسلة OpenAI.
+- 01-08-2025 يدعم سير العمل الوكيل وMCP.
+- 23-05-2025 تمت إضافة مكون منفذ كود Python/JavaScript إلى Agent.
+- 05-05-2025 يدعم الاستعلام بين اللغات.
+- 19-03-2025 يدعم استخدام نموذج متعدد الوسائط لفهم الصور داخل ملفات PDF أو DOCX.
+
+## 🎉 تابعونا
+
+⭐️ قم بتمييز مستودعنا بنجمة لتبقى على اطلاع بالميزات والتحسينات الجديدة والمثيرة! احصل على إشعارات فورية بالجديد
+الإصدارات! 🌟
+
+<div align="center" style="margin-top:20px;margin-bottom:20px;">
+<img src="https://github.com/user-attachments/assets/18c9707e-b8aa-4caf-a154-037089c105ba" width="1200"/>
+</div>
+
+## 🌟 الميزات الرئيسية
+
+### 🍭 **"الجودة في الداخل، الجودة في الخارج"**
+
+- [الفهم العميق للمستندات](./deepdoc/README.md) لاستخراج المعرفة من البيانات غير المنظمة
+  ذات التنسيقات المعقدة.
+- يجد "إبرة في كومة قش بيانات" من الرموز غير المحدودة حرفيًا.
+
+### 🍱 **التقطيع القائم على القالب**
+
+- ذكي وقابل للتفسير.
+- الكثير من خيارات القالب للاختيار من بينها.
+
+### 🌱 **استشهادات مؤرضة لتقليل الهلوسة**
+
+- تصور تقطيع النص للسماح بالتدخل البشري.
+- عرض سريع للمراجع الرئيسية والاستشهادات التي يمكن تتبعها لدعم الإجابات المبنية على أسس سليمة.
+
+### 🍔 **التوافق مع مصادر البيانات غير المتجانسة**
+
+- يدعم Word، والشرائح، وExcel، وtxt، والصور، والنسخ الممسوحة ضوئيًا، والبيانات المنظمة، وصفحات الويب، والمزيد.
+
+### 🛀 **سير عمل RAG آلي وسهل**
+
+- تنسيق RAG مبسط يلبي احتياجات الشركات الشخصية والكبيرة على حد سواء.
+- نماذج LLMs قابلة للتكوين بالإضافة إلى نماذج embedding.
+- الاستدعاء المتعدد المقترن بإعادة التصنيف المدمجة.
+- APIs بديهي للتكامل السلس مع الأعمال.
+
+## 🔎 هندسة النظام
+
+<div align="center" style="margin-top:20px;margin-bottom:20px;">
+<img src="https://github.com/user-attachments/assets/31b0dd6f-ca4f-445a-9457-70cb44a381b2" width="1000"/>
+</div>
+
+## 🎬 ابدأ
+
+### 📝 المتطلبات الأساسية
+
+- CPU >= 4 مراكز
+- الرام >= 16 جيجا
+- القرص >= 50 جيجا بايت
+- Docker >= 24.0.0 & Docker Compose >= v2.26.1
+- [gVisor](https://gvisor.dev/docs/user_guide/install/): مطلوب فقط إذا كنت تنوي استخدام ميزة منفذ التعليمات البرمجية (وضع الحماية) لـ RAGFlow.
+
+> [!TIP]
+> إذا لم تقم بتثبيت Docker على جهازك المحلي (Windows أو Mac أو Linux)، راجع [تثبيت Docker Engine](https://docs.docker.com/engine/install/).
+
+### 🚀 بدء تشغيل الخادم
+
+1. تأكد من `vm.max_map_count` >= 262144:
+
+   > للتحقق من قيمة `vm.max_map_count`:
+   >
+   > ```bash
+   > $ sysctl vm.max_map_count
+   > ```
+   >
+   > أعد تعيين `vm.max_map_count` إلى قيمة 262144 على الأقل إذا لم تكن كذلك.
+   >
+   > ```bash
+   > # In this case, we set it to 262144:
+   > $ sudo sysctl -w vm.max_map_count=262144
+   > ```
+   >
+   > سيتم إعادة ضبط هذا التغيير بعد إعادة تشغيل النظام. لضمان بقاء التغيير دائمًا، قم بإضافة أو تحديث
+   > `vm.max_map_count` القيمة في **/etc/sysctl.conf** وفقًا لذلك:
+   >
+   > ```bash
+   > vm.max_map_count=262144
+   > ```
+   >
+2. استنساخ الريبو:
+
+   ```bash
+   $ git clone https://github.com/infiniflow/ragflow.git
+   ```
+3. ابدأ تشغيل الخادم باستخدام صور Docker المعدة مسبقًا:
+
+> [!CAUTION]
+> جميع الصور Docker مصممة لمنصات x86. لا نعرض حاليًا صور Docker لـ ARM64.
+> إذا كنت تستخدم نظامًا أساسيًا ARM64، فاتبع [هذا الدليل](https://ragflow.io/docs/dev/build_docker_image) لإنشاء صورة Docker متوافقة مع نظامك.
+
+> يقوم الأمر أدناه بتنزيل إصدار `v0.25.0` من الصورة RAGFlow Docker. راجع الجدول التالي للحصول على أوصاف لإصدارات RAGFlow المختلفة. لتنزيل إصدار RAGFlow مختلف عن `v0.25.0`، قم بتحديث المتغير `RAGFLOW_IMAGE` وفقًا لذلك في **docker/.env** قبل استخدام `docker compose` لبدء تشغيل الخادم.
+
+```bash
+   $ cd ragflow/docker
+
+   # git checkout v0.25.0
+   # Optional: use a stable tag (see releases: https://github.com/infiniflow/ragflow/releases)
+   # This step ensures the **entrypoint.sh** file in the code matches the Docker image version.
+
+   # Use CPU for DeepDoc tasks:
+   $ docker compose -f docker-compose.yml up -d
+
+   # To use GPU to accelerate DeepDoc tasks:
+   # sed -i '1i DEVICE=gpu' .env
+   # docker compose -f docker-compose.yml up -d
+```
+
+> ملاحظة: قبل `v0.22.0`، قدمنا ​​كلتا الصورتين بنماذج embedding وصورًا رفيعة بدون نماذج embedding. التفاصيل على النحو التالي:
+
+| RAGFlow علامة الصورة | حجم الصورة (جيجابايت) | هل لديه نماذج embedding؟ | مستقر؟        |
+|-------------------|-----------------|-----------------------|----------------|
+| v0.21.1 | &approx;9 | ✔️ | إصدار مستقر |
+| v0.21.1-slim | &approx;2 | ❌ | إصدار مستقر |
+
+> بدءًا من `v0.22.0`، نقوم بشحن الإصدار النحيف فقط ولم نعد نلحق اللاحقة **-slim** بعلامة الصورة.
+
+4. التحقق من حالة الخادم بعد تشغيل الخادم:
+
+   ```bash
+   $ docker logs -f docker-ragflow-cpu-1
+   ```
+
+   _النتيجة التالية تؤكد الإطلاق الناجح للنظام:_
+
+   ```bash
+
+         ____   ___    ______ ______ __
+        / __ \ /   |  / ____// ____// /____  _      __
+       / /_/ // /| | / / __ / /_   / // __ \| | /| / /
+      / _, _// ___ |/ /_/ // __/  / // /_/ /| |/ |/ /
+     /_/ |_|/_/  |_|\____//_/    /_/ \____/ |__/|__/
+
+    * Running on all addresses (0.0.0.0)
+   ```
+
+   > إذا تخطيت خطوة التأكيد هذه وقمت بتسجيل الدخول مباشرة إلى RAGFlow، فقد يعرض متصفحك تنبيه `network abnormal`
+   > خطأ لأنه في تلك اللحظة، قد لا تتم تهيئة RAGFlow بشكل كامل.
+   >
+5. في متصفح الويب الخاص بك، أدخل عنوان IP الخاص بالخادم الخاص بك وقم بتسجيل الدخول إلى RAGFlow.
+
+   > باستخدام الإعدادات الافتراضية، ما عليك سوى إدخال `http://IP_OF_YOUR_MACHINE` (**من دون** رقم المنفذ) كإعداد افتراضي
+   > HTTP يمكن حذف منفذ العرض `80` عند استخدام التكوينات الافتراضية.
+   >
+6. في [service_conf.yaml.template](./docker/service_conf.yaml.template)، حدد المصنع LLM المطلوب في `user_default_llm` وقم بالتحديث
+   الحقل `API_KEY` مع مفتاح API المقابل.
+
+   > راجع [llm_api_key_setup](https://ragflow.io/docs/dev/llm_api_key_setup) لمزيد من المعلومات.
+   >
+
+   _العرض بدأ!_
+
+## 🔧 التكوينات
+
+عندما يتعلق الأمر بتكوينات النظام، ستحتاج إلى إدارة الملفات التالية:
+
+- [.env](./docker/.env): يحتفظ بالإعدادات الأساسية للنظام، مثل `SVR_HTTP_PORT`، `MYSQL_PASSWORD`، و
+  `MINIO_PASSWORD`.
+- [service_conf.yaml.template](./docker/service_conf.yaml.template): تكوين الخدمات الخلفية. سيتم ملء متغيرات البيئة في هذا الملف تلقائيًا عند بدء تشغيل الحاوية Docker. ستكون أي متغيرات بيئة تم تعيينها داخل حاوية Docker متاحة للاستخدام، مما يسمح لك بتخصيص سلوك الخدمة استنادًا إلى بيئة النشر.
+- [docker-compose.yml](./docker/docker-compose.yml): يعتمد النظام على [docker-compose.yml](./docker/docker-compose.yml) لبدء التشغيل.
+
+> يوفر الملف [./docker/README](./docker/README.md) وصفًا تفصيليًا لإعدادات البيئة والخدمة
+> التكوينات التي يمكن استخدامها كـ `${ENV_VARS}` في ملف [service_conf.yaml.template](./docker/service_conf.yaml.template).
+
+لتحديث منفذ العرض الافتراضي HTTP (80)، انتقل إلى [docker-compose.yml](./docker/docker-compose.yml) وقم بتغيير `80:80`
+إلى `<YOUR_SERVING_PORT>:80`.
+
+تتطلب تحديثات التكوينات المذكورة أعلاه إعادة تشغيل جميع الحاويات لتصبح سارية المفعول:
+
+> ```bash
+> $ docker compose -f docker-compose.yml up -d
+> ```
+
+### تبديل محرك المستندات من Elasticsearch إلى Infinity
+
+RAGFlow يستخدم Elasticsearch بشكل افتراضي لتخزين النص الكامل والمتجهات. للتبديل إلى [Infinity](https://github.com/infiniflow/infinity/)، اتبع الخطوات التالية:
+
+1. إيقاف كافة الحاويات قيد التشغيل:
+
+   ```bash
+   $ docker compose -f docker/docker-compose.yml down -v
+   ```
+
+> [!WARNING]
+> `-v` سوف يحذف docker وحدات تخزين الحاوية، وسيتم مسح البيانات الموجودة.
+
+2. اضبط `DOC_ENGINE` في **docker/.env** على `infinity`.
+3. ابدأ الحاويات:
+
+   ```bash
+   $ docker compose -f docker-compose.yml up -d
+   ```
+
+> [!WARNING]
+> التبديل إلى Infinity على جهاز Linux/arm64 غير مدعوم رسميًا بعد.
+
+## 🔧 أنشئ صورة Docker
+
+يبلغ حجم هذه الصورة حوالي 2 غيغابايت وتعتمد على خدمات LLM وembedding الخارجية.
+
+```bash
+git clone https://github.com/infiniflow/ragflow.git
+cd ragflow/
+docker build --platform linux/amd64 -f Dockerfile -t infiniflow/ragflow:nightly .
+```
+
+أو إذا كنت خلف وكيل، فيمكنك تمرير وسيطات الوكيل:
+
+```bash
+docker build --platform linux/amd64 \
+  --build-arg http_proxy=http://YOUR_PROXY:PORT \
+  --build-arg https_proxy=http://YOUR_PROXY:PORT \
+  -f Dockerfile -t infiniflow/ragflow:nightly .
+```
+
+## 🔨 إطلاق الخدمة من المصدر للتطوير
+
+1. قم بتثبيت `uv` و`pre-commit`، أو قم بتخطي هذه الخطوة إذا كانا مثبتين بالفعل:
+
+   ```bash
+   pipx install uv pre-commit
+   ```
+2. استنساخ الكود المصدري وتثبيت تبعيات بايثون:
+
+   ```bash
+   git clone https://github.com/infiniflow/ragflow.git
+   cd ragflow/
+   uv sync --python 3.12 # install RAGFlow dependent python modules
+   uv run python3 download_deps.py
+   pre-commit install
+   ```
+3. قم بتشغيل الخدمات التابعة (MinIO وElasticsearch وRedis وMySQL) باستخدام Docker Compose:
+
+   ```bash
+   docker compose -f docker/docker-compose-base.yml up -d
+   ```
+
+   أضف السطر التالي إلى `/etc/hosts` لحل كافة المضيفين المحددين في **docker/.env** إلى `127.0.0.1`:
+
+   ```
+   127.0.0.1       es01 infinity mysql minio redis sandbox-executor-manager
+   ```
+4. إذا لم تتمكن من الوصول إلى HuggingFace، فقم بتعيين متغير البيئة `HF_ENDPOINT` لاستخدام موقع مرآة:
+
+   ```bash
+   export HF_ENDPOINT=https://hf-mirror.com
+   ```
+5. إذا كان نظام التشغيل لديك لا يحتوي على jemalloc، فيرجى تثبيته على النحو التالي:
+
+   ```bash
+   # Ubuntu
+   sudo apt-get install libjemalloc-dev
+   # CentOS
+   sudo yum install jemalloc
+   # OpenSUSE
+   sudo zypper install jemalloc
+   # macOS
+   sudo brew install jemalloc
+   ```
+6. إطلاق الخدمة الخلفية:
+
+   ```bash
+   source .venv/bin/activate
+   export PYTHONPATH=$(pwd)
+   bash docker/launch_backend_service.sh
+   ```
+7. تثبيت تبعيات الواجهة الأمامية:
+
+   ```bash
+   cd web
+   npm install
+   ```
+8. إطلاق خدمة الواجهة الأمامية:
+
+   ```bash
+   npm run dev
+   ```
+
+   _النتيجة التالية تؤكد الإطلاق الناجح للنظام:_
+
+   ![](https://github.com/user-attachments/assets/0daf462c-a24d-4496-a66f-92533534e187)
+9. أوقف خدمة الواجهة الأمامية والخلفية RAGFlow بعد اكتمال التطوير:
+
+   ```bash
+   pkill -f "ragflow_server.py|task_executor.py"
+   ```
+
+## 📚 التوثيق
+
+- [البدء السريع](https://ragflow.io/docs/dev/)
+- [التكوين](https://ragflow.io/docs/dev/configurations)
+- [ملاحظات الإصدار](https://ragflow.io/docs/dev/release_notes)
+- [أدلة المستخدم](https://ragflow.io/docs/category/user-guides)
+- [أدلة المطورين](https://ragflow.io/docs/category/developer-guides)
+- [المراجع](https://ragflow.io/docs/dev/category/references)
+- [الأسئلة الشائعة](https://ragflow.io/docs/dev/faq)
+
+## 📜 Roadmap
+
+راجع [RAGFlow Roadmap 2026](https://github.com/infiniflow/ragflow/issues/12241)
+
+## 🏄 المجتمع
+
+- [Discord](https://discord.gg/NjYzJD3GM3)
+- [Twitter](https://twitter.com/infiniflowai)
+- [مناقشات جيثب](https://github.com/orgs/infiniflow/discussions)
+
+## 🙌 المساهمة
+
+RAGFlow يزدهر من خلال التعاون مفتوح المصدر. وبهذه الروح، فإننا نحتضن المساهمات المتنوعة من المجتمع.
+إذا كنت ترغب في أن تكون جزءًا، فراجع [إرشادات المساهمة](https://ragflow.io/docs/dev/contributing) أولاً.

README_fr.md

@@ -0,0 +1,405 @@
+<div align="center">
+<a href="https://cloud.ragflow.io/">
+<img src="web/src/assets/logo-with-text.svg" width="520" alt="ragflow logo">
+</a>
+</div>
+
+<p align="center">
+  <a href="./README.md"><img alt="README in English" src="https://img.shields.io/badge/English-DFE0E5"></a>
+  <a href="./README_zh.md"><img alt="简体中文版自述文件" src="https://img.shields.io/badge/简体中文-DFE0E5"></a>
+  <a href="./README_tzh.md"><img alt="繁體版中文自述文件" src="https://img.shields.io/badge/繁體中文-DFE0E5"></a>
+  <a href="./README_ja.md"><img alt="日本語のREADME" src="https://img.shields.io/badge/日本語-DFE0E5"></a>
+  <a href="./README_ko.md"><img alt="한국어" src="https://img.shields.io/badge/한국어-DFE0E5"></a>
+  <a href="./README_id.md"><img alt="Bahasa Indonesia" src="https://img.shields.io/badge/Bahasa Indonesia-DFE0E5"></a>
+  <a href="./README_pt_br.md"><img alt="Português(Brasil)" src="https://img.shields.io/badge/Português(Brasil)-DFE0E5"></a>
+  <a href="./README_fr.md"><img alt="README en Français" src="https://img.shields.io/badge/Français-DBEDFA"></a>
+  <a href="./README_ar.md"><img alt="README in Arabic" src="https://img.shields.io/badge/Arabic-DFE0E5"></a>
+  <a href="./README_tr.md"><img alt="Türkçe README" src="https://img.shields.io/badge/Türkçe-DFE0E5"></a>
+</p>
+
+<p align="center">
+    <a href="https://x.com/intent/follow?screen_name=infiniflowai" target="_blank">
+        <img src="https://img.shields.io/twitter/follow/infiniflow?logo=X&color=%20%23f5f5f5" alt="suivre sur X(Twitter)">
+    </a>
+    <a href="https://cloud.ragflow.io" target="_blank">
+        <img alt="Badge statique" src="https://img.shields.io/badge/Online-Demo-4e6b99">
+    </a>
+    <a href="https://hub.docker.com/r/infiniflow/ragflow" target="_blank">
+        <img src="https://img.shields.io/docker/pulls/infiniflow/ragflow?label=Docker%20Pulls&color=0db7ed&logo=docker&logoColor=white&style=flat-square" alt="docker pull infiniflow/ragflow:v0.25.0">
+    </a>
+    <a href="https://github.com/infiniflow/ragflow/releases/latest">
+        <img src="https://img.shields.io/github/v/release/infiniflow/ragflow?color=blue&label=Dernière%20version" alt="Dernière version">
+    </a>
+    <a href="https://github.com/infiniflow/ragflow/blob/main/LICENSE">
+        <img height="21" src="https://img.shields.io/badge/License-Apache--2.0-ffffff?labelColor=d4eaf7&color=2e6cc4" alt="licence">
+    </a>
+    <a href="https://deepwiki.com/infiniflow/ragflow">
+        <img alt="Ask DeepWiki" src="https://deepwiki.com/badge.svg">
+    </a>
+</p>
+
+<h4 align="center">
+  <a href="https://ragflow.io/docs/dev/">Documentation</a> |
+  <a href="https://github.com/infiniflow/ragflow/issues/12241">Roadmap</a> |
+  <a href="https://twitter.com/infiniflowai">Twitter</a> |
+  <a href="https://discord.gg/NjYzJD3GM3">Discord</a> |
+  <a href="https://cloud.ragflow.io">Démo</a>
+</h4>
+
+<div align="center" style="margin-top:20px;margin-bottom:20px;">
+<img src="https://raw.githubusercontent.com/infiniflow/ragflow-docs/refs/heads/image/image/ragflow-octoverse.png" width="1200"/>
+</div>
+
+<div align="center">
+<a href="https://trendshift.io/repositories/9064" target="_blank"><img src="https://trendshift.io/api/badge/repositories/9064" alt="infiniflow%2Fragflow | Trendshift" style="width: 250px; height: 55px;" width="250" height="55"/></a>
+</div>
+
+<details open>
+<summary><b>📕 Table des matières</b></summary>
+
+- 💡 [Qu'est-ce que RAGFlow?](#-quest-ce-que-ragflow)
+- 🎮 [Démo](#-démo)
+- 📌 [Dernières mises à jour](#-dernières-mises-à-jour)
+- 🌟 [Fonctionnalités clés](#-fonctionnalités-clés)
+- 🔎 [Architecture du système](#-architecture-du-système)
+- 🎬 [Démarrage](#-démarrage)
+- 🔧 [Configurations](#-configurations)
+- 🔧 [Construire une image Docker](#-construire-une-image-docker)
+- 🔨 [Lancer le service depuis les sources pour le développement](#-lancer-le-service-depuis-les-sources-pour-le-développement)
+- 📚 [Documentation](#-documentation)
+- 📜 [Roadmap](#-feuille-de-route)
+- 🏄 [Communauté](#-communauté)
+- 🙌 [Contribuer](#-contribuer)
+
+</details>
+
+## 💡 Qu'est-ce que RAGFlow?
+
+[RAGFlow](https://ragflow.io/) est un moteur de [RAG](https://ragflow.io/basics/what-is-rag) (Retrieval-Augmented Generation) open-source de premier plan qui fusionne les technologies RAG de pointe avec des capacités Agent pour créer une couche de contexte supérieure pour les LLM. Il offre un flux de travail RAG rationalisé, adaptable aux entreprises de toute taille. Alimenté par un [moteur de contexte](https://ragflow.io/basics/what-is-agent-context-engine) convergent et des modèles d'agents préconstruits, RAGFlow permet aux développeurs de transformer des données complexes en systèmes d'IA haute-fidélité, prêts pour la production, avec une efficacité et une précision exceptionnelles.
+
+## 🎮 Démo
+
+Essayez notre démo sur [https://cloud.ragflow.io](https://cloud.ragflow.io).
+
+<div align="center" style="margin-top:20px;margin-bottom:20px;">
+<img src="https://raw.githubusercontent.com/infiniflow/ragflow-docs/refs/heads/image/image/chunking.gif" width="1200"/>
+<img src="https://raw.githubusercontent.com/infiniflow/ragflow-docs/refs/heads/image/image/agentic-dark.gif" width="1200"/>
+</div>
+
+## 🔥 Dernières mises à jour
+
+- 24-03-2026 [RAGFlow Skill on OpenClaw](https://clawhub.ai/yingfeng/ragflow-skill) — Fournit un skill officiel pour accéder aux datasets RAGFlow via OpenClaw.
+- 26-12-2025 Prise en charge de la « Mémoire » pour l'agent IA.
+- 19-11-2025 Prise en charge de Gemini 3 Pro.
+- 12-11-2025 Prise en charge de la synchronisation de données depuis Confluence, S3, Notion, Discord et Google Drive.
+- 23-10-2025 Prise en charge de MinerU & Docling comme méthodes d'analyse de documents.
+- 15-10-2025 Prise en charge du pipeline d'ingestion orchestrable.
+- 08-08-2025 Prise en charge des derniers modèles de la série GPT-5 d'OpenAI.
+- 01-08-2025 Prise en charge du flux de travail agentique et de MCP.
+- 23-05-2025 Ajout d'un composant exécuteur de code Python/JavaScript à l'Agent.
+- 05-05-2025 Prise en charge des requêtes inter-langues.
+- 19-03-2025 Prise en charge de l'utilisation d'un modèle multi-modal pour analyser les images dans les fichiers PDF ou DOCX.
+
+## 🎉 Restez informé
+
+⭐️ Mettez une étoile à notre dépôt pour rester informé des nouvelles fonctionnalités et améliorations passionnantes ! Recevez des notifications instantanées pour les nouvelles versions ! 🌟
+
+<div align="center" style="margin-top:20px;margin-bottom:20px;">
+<img src="https://github.com/user-attachments/assets/18c9707e-b8aa-4caf-a154-037089c105ba" width="1200"/>
+</div>
+
+## 🌟 Fonctionnalités clés
+
+### 🍭 **"Quality in, quality out"**
+
+- Extraction de connaissances basée sur la [compréhension approfondie des documents](./deepdoc/README.md) à partir de données non structurées aux formats complexes.
+- Trouve "l'aiguille dans la meule de données" de tokens littéralement illimités.
+
+### 🍱 **Découpage(Chunking) basé sur des templates**
+
+- Intelligent et explicable.
+- De nombreuses options de templates disponibles.
+
+### 🌱 **Citations fondées avec réduction des hallucinations**
+
+- Visualisation du découpage de texte pour permettre une intervention humaine.
+- Aperçu rapide des références clés et citations traçables pour soutenir des réponses fondées.
+
+### 🍔 **Compatibilité avec des sources de données hétérogènes**
+
+- Prend en charge Word, présentations, Excel, txt, images, copies numérisées, données structurées, pages web, et plus encore.
+
+### 🛀 **Flux de travail RAG automatisé et sans effort**
+
+- Orchestration RAG rationalisée adaptée aux particuliers comme aux grandes entreprises.
+- LLM et modèles d'embedding configurables.
+- Rappel multiple associé à un ré-classement fusionné.
+- APIs intuitives pour une intégration transparente avec les entreprises.
+
+## 🔎 Architecture du système
+
+<div align="center" style="margin-top:20px;margin-bottom:20px;">
+<img src="https://github.com/user-attachments/assets/31b0dd6f-ca4f-445a-9457-70cb44a381b2" width="1000"/>
+</div>
+
+## 🎬 Démarrage
+
+### 📝 Prérequis
+
+- CPU >= 4 cœurs
+- RAM >= 16 Go
+- Disque >= 50 Go
+- Docker >= 24.0.0 & Docker Compose >= v2.26.1
+- [gVisor](https://gvisor.dev/docs/user_guide/install/) : Requis uniquement si vous souhaitez utiliser la fonctionnalité d'exécuteur de code (sandbox) de RAGFlow.
+
+> [!TIP]
+> Si vous n'avez pas installé Docker sur votre machine locale (Windows, Mac ou Linux), consultez [Installer Docker Engine](https://docs.docker.com/engine/install/).
+
+### 🚀 Démarrer le serveur
+
+1. Assurez-vous que `vm.max_map_count` >= 262144 :
+
+   > Pour vérifier la valeur de `vm.max_map_count` :
+   >
+   > ```bash
+   > $ sysctl vm.max_map_count
+   > ```
+   >
+   > Réinitialisez `vm.max_map_count` à une valeur d'au moins 262144 si ce n'est pas le cas.
+   >
+   > ```bash
+   > # Dans ce cas, nous le définissons à 262144 :
+   > $ sudo sysctl -w vm.max_map_count=262144
+   > ```
+   >
+   > Ce changement sera réinitialisé après un redémarrage du système. Pour que votre modification reste permanente, ajoutez ou mettez à jour la valeur `vm.max_map_count` dans **/etc/sysctl.conf** :
+   >
+   > ```bash
+   > vm.max_map_count=262144
+   > ```
+   >
+2. Clonez le dépôt :
+
+   ```bash
+   $ git clone https://github.com/infiniflow/ragflow.git
+   ```
+3. Démarrez le serveur en utilisant les images Docker préconstruites :
+
+> [!CAUTION]
+> Toutes les images Docker sont construites pour les plateformes x86. Nous ne proposons pas actuellement d'images Docker pour ARM64.
+> Si vous êtes sur une plateforme ARM64, suivez [ce guide](https://ragflow.io/docs/dev/build_docker_image) pour construire une image Docker compatible avec votre système.
+
+> La commande ci-dessous télécharge l'édition `v0.25.0` de l'image Docker RAGFlow. Consultez le tableau suivant pour les descriptions des différentes éditions de RAGFlow. Pour télécharger une édition de RAGFlow différente de `v0.25.0`, mettez à jour la variable `RAGFLOW_IMAGE` dans **docker/.env** avant d'utiliser `docker compose` pour démarrer le serveur.
+
+```bash
+   $ cd ragflow/docker
+
+   # git checkout v0.25.0
+   # Optionnel : utiliser un tag stable (voir les versions : https://github.com/infiniflow/ragflow/releases)
+   # Cette étape garantit que le fichier **entrypoint.sh** dans le code correspond à la version de l'image Docker.
+
+   # Use CPU for DeepDoc tasks:
+   $ docker compose -f docker-compose.yml up -d
+
+   # To use GPU to accelerate DeepDoc tasks:
+   # sed -i '1i DEVICE=gpu' .env
+   # docker compose -f docker-compose.yml up -d
+```
+
+> Remarque : Avant `v0.22.0`, nous fournissions à la fois des images avec des modèles d'embedding et des images slim sans modèles d'embedding. Détails ci-dessous :
+
+| RAGFlow image tag | Image size (GB) | Has embedding models? | Stable?        |
+|-------------------|-----------------|-----------------------|----------------|
+| v0.21.1           | &approx;9       | ✔️                    | Stable release |
+| v0.21.1-slim      | &approx;2       | ❌                     | Stable release |
+
+> À partir de `v0.22.0`, nous ne distribuons que l'édition slim et ne rajoutons plus le suffixe **-slim** au tag d'image.
+
+4. Vérifiez l'état du serveur après son démarrage :
+
+   ```bash
+   $ docker logs -f docker-ragflow-cpu-1
+   ```
+
+   _La sortie suivante confirme un lancement réussi du système :_
+
+   ```bash
+
+         ____   ___    ______ ______ __
+        / __ \ /   |  / ____// ____// /____  _      __
+       / /_/ // /| | / / __ / /_   / // __ \| | /| / /
+      / _, _// ___ |/ /_/ // __/  / // /_/ /| |/ |/ /
+     /_/ |_|/_/  |_|\____//_/    /_/ \____/ |__/|__/
+
+    * Running on all addresses (0.0.0.0)
+   ```
+
+   > Si vous sautez cette étape de confirmation et vous connectez directement à RAGFlow, votre navigateur peut afficher une erreur `network abnormal`, car à ce moment-là, votre RAGFlow peut ne pas être entièrement initialisé.
+   >
+5. Dans votre navigateur web, entrez l'adresse IP de votre serveur et connectez-vous à RAGFlow.
+
+   > Avec les paramètres par défaut, il vous suffit d'entrer `http://IP_OF_YOUR_MACHINE` (**sans** numéro de port), car le port HTTP par défaut `80` peut être omis lors de l'utilisation des configurations par défaut.
+   >
+6. Dans [service_conf.yaml.template](./docker/service_conf.yaml.template), sélectionnez la fabrique LLM souhaitée dans `user_default_llm` et mettez à jour le champ `API_KEY` avec la clé API correspondante.
+
+   > Voir [llm_api_key_setup](https://ragflow.io/docs/dev/llm_api_key_setup) pour plus d'informations.
+   >
+
+   _Le spectacle commence !_
+
+## 🔧 Configurations
+
+En ce qui concerne les configurations système, vous devrez gérer les fichiers suivants :
+
+- [.env](./docker/.env) : Conserve les paramètres de base du système, tels que `SVR_HTTP_PORT`, `MYSQL_PASSWORD` et `MINIO_PASSWORD`.
+- [service_conf.yaml.template](./docker/service_conf.yaml.template) : Configure les services back-end. Les variables d'environnement dans ce fichier seront automatiquement renseignées au démarrage du conteneur Docker. Toutes les variables d'environnement définies dans le conteneur Docker seront disponibles, vous permettant de personnaliser le comportement du service en fonction de l'environnement de déploiement.
+- [docker-compose.yml](./docker/docker-compose.yml) : Le système s'appuie sur [docker-compose.yml](./docker/docker-compose.yml) pour démarrer.
+
+> Le fichier [./docker/README](./docker/README.md) fournit une description détaillée des paramètres d'environnement et des configurations de services qui peuvent être utilisés comme `${ENV_VARS}` dans le fichier [service_conf.yaml.template](./docker/service_conf.yaml.template).
+
+Pour mettre à jour le port HTTP de service par défaut (80), accédez à [docker-compose.yml](./docker/docker-compose.yml) et changez `80:80` en `<YOUR_SERVING_PORT>:80`.
+
+Les mises à jour des configurations ci-dessus nécessitent un redémarrage de tous les conteneurs pour prendre effet :
+
+> ```bash
+> $ docker compose -f docker-compose.yml up -d
+> ```
+
+### Passer du moteur de documents Elasticsearch à Infinity
+
+RAGFlow utilise Elasticsearch par défaut pour stocker le texte intégral et les vecteurs. Pour passer à [Infinity](https://github.com/infiniflow/infinity/), suivez ces étapes :
+
+1. Arrêtez tous les conteneurs en cours d'exécution :
+
+   ```bash
+   $ docker compose -f docker/docker-compose.yml down -v
+   ```
+
+> [!WARNING]
+> `-v` supprimera les volumes des conteneurs Docker, et les données existantes seront effacées.
+
+2. Définissez `DOC_ENGINE` dans **docker/.env** sur `infinity`.
+3. Démarrez les conteneurs :
+
+   ```bash
+   $ docker compose -f docker-compose.yml up -d
+   ```
+
+> [!WARNING]
+> Le passage à Infinity sur une machine Linux/arm64 n'est pas encore officiellement pris en charge.
+
+## 🔧 Construire une image Docker
+
+Cette image fait environ 2 Go et dépend de services LLM et d'embedding externes.
+
+```bash
+git clone https://github.com/infiniflow/ragflow.git
+cd ragflow/
+docker build --platform linux/amd64 -f Dockerfile -t infiniflow/ragflow:nightly .
+```
+
+Ou si vous êtes derrière un proxy, vous pouvez passer des arguments de proxy :
+
+```bash
+docker build --platform linux/amd64 \
+  --build-arg http_proxy=http://YOUR_PROXY:PORT \
+  --build-arg https_proxy=http://YOUR_PROXY:PORT \
+  -f Dockerfile -t infiniflow/ragflow:nightly .
+```
+
+## 🔨 Lancer le service depuis les sources pour le développement
+
+1. Installez `uv` et `pre-commit`, ou ignorez cette étape s'ils sont déjà installés :
+
+   ```bash
+   pipx install uv pre-commit
+   ```
+2. Clonez le code source et installez les dépendances Python :
+
+   ```bash
+   git clone https://github.com/infiniflow/ragflow.git
+   cd ragflow/
+   uv sync --python 3.12 # install RAGFlow dependent python modules
+   uv run python3 download_deps.py
+   pre-commit install
+   ```
+3. Lancez les services dépendants (MinIO, Elasticsearch, Redis et MySQL) avec Docker Compose :
+
+   ```bash
+   docker compose -f docker/docker-compose-base.yml up -d
+   ```
+
+   Ajoutez la ligne suivante à `/etc/hosts` pour résoudre tous les hôtes spécifiés dans **docker/.env** vers `127.0.0.1` :
+
+   ```
+   127.0.0.1       es01 infinity mysql minio redis sandbox-executor-manager
+   ```
+4. Si vous ne pouvez pas accéder à HuggingFace, définissez la variable d'environnement `HF_ENDPOINT` pour utiliser un site miroir :
+
+   ```bash
+   export HF_ENDPOINT=https://hf-mirror.com
+   ```
+5. Si votre système d'exploitation n'a pas jemalloc, installez-le comme suit :
+
+   ```bash
+   # Ubuntu
+   sudo apt-get install libjemalloc-dev
+   # CentOS
+   sudo yum install jemalloc
+   # OpenSUSE
+   sudo zypper install jemalloc
+   # macOS
+   sudo brew install jemalloc
+   ```
+6. Lancez le service back-end :
+
+   ```bash
+   source .venv/bin/activate
+   export PYTHONPATH=$(pwd)
+   bash docker/launch_backend_service.sh
+   ```
+7. Installez les dépendances front-end :
+
+   ```bash
+   cd web
+   npm install
+   ```
+8. Lancez le service front-end :
+
+   ```bash
+   npm run dev
+   ```
+
+   _La sortie suivante confirme un lancement réussi du système :_
+
+   ![](https://github.com/user-attachments/assets/0daf462c-a24d-4496-a66f-92533534e187)
+9. Arrêtez les services front-end et back-end de RAGFlow une fois le développement terminé :
+
+   ```bash
+   pkill -f "ragflow_server.py|task_executor.py"
+   ```
+
+## 📚 Documentation
+
+- [Quickstart](https://ragflow.io/docs/dev/)
+- [Configuration](https://ragflow.io/docs/dev/configurations)
+- [Release notes](https://ragflow.io/docs/dev/release_notes)
+- [User guides](https://ragflow.io/docs/category/user-guides)
+- [Developer guides](https://ragflow.io/docs/category/developer-guides)
+- [References](https://ragflow.io/docs/dev/category/references)
+- [FAQs](https://ragflow.io/docs/dev/faq)
+
+## 📜 Roadmap
+
+Voir la [Feuille de route RAGFlow 2026](https://github.com/infiniflow/ragflow/issues/12241)
+
+## 🏄 Communauté
+
+- [Discord](https://discord.gg/NjYzJD3GM3)
+- [Twitter](https://twitter.com/infiniflowai)
+- [GitHub Discussions](https://github.com/orgs/infiniflow/discussions)
+
+## 🙌 Contribuer
+
+RAGFlow s'épanouit grâce à la collaboration open-source. Dans cet esprit, nous accueillons des contributions diverses de la communauté.
+Si vous souhaitez en faire partie, consultez d'abord nos [Directives de contribution](https://ragflow.io/docs/dev/contributing).

README_id.md

@@ -1,5 +1,5 @@
 <div align="center">
-<a href="https://demo.ragflow.io/">
+<a href="https://cloud.ragflow.io/">
 <img src="web/src/assets/logo-with-text.svg" width="520" alt="Logo ragflow">
 </a>
 </div>
@@ -12,17 +12,20 @@
   <a href="./README_ko.md"><img alt="한국어" src="https://img.shields.io/badge/한국어-DFE0E5"></a>
   <a href="./README_id.md"><img alt="Bahasa Indonesia" src="https://img.shields.io/badge/Bahasa Indonesia-DBEDFA"></a>
   <a href="./README_pt_br.md"><img alt="Português(Brasil)" src="https://img.shields.io/badge/Português(Brasil)-DFE0E5"></a>
+  <a href="./README_fr.md"><img alt="README en Français" src="https://img.shields.io/badge/Français-DFE0E5"></a>
+  <a href="./README_ar.md"><img alt="README in Arabic" src="https://img.shields.io/badge/Arabic-DFE0E5"></a>
+  <a href="./README_tr.md"><img alt="Türkçe README" src="https://img.shields.io/badge/Türkçe-DFE0E5"></a>
 </p>
 
 <p align="center">
     <a href="https://x.com/intent/follow?screen_name=infiniflowai" target="_blank">
         <img src="https://img.shields.io/twitter/follow/infiniflow?logo=X&color=%20%23f5f5f5" alt="Ikuti di X (Twitter)">
     </a>
-    <a href="https://demo.ragflow.io" target="_blank">
+    <a href="https://cloud.ragflow.io" target="_blank">
         <img alt="Lencana Daring" src="https://img.shields.io/badge/Online-Demo-4e6b99">
     </a>
     <a href="https://hub.docker.com/r/infiniflow/ragflow" target="_blank">
-        <img src="https://img.shields.io/docker/pulls/infiniflow/ragflow?label=Docker%20Pulls&color=0db7ed&logo=docker&logoColor=white&style=flat-square" alt="docker pull infiniflow/ragflow:v0.24.0">
+        <img src="https://img.shields.io/docker/pulls/infiniflow/ragflow?label=Docker%20Pulls&color=0db7ed&logo=docker&logoColor=white&style=flat-square" alt="docker pull infiniflow/ragflow:v0.25.0">
     </a>
     <a href="https://github.com/infiniflow/ragflow/releases/latest">
         <img src="https://img.shields.io/github/v/release/infiniflow/ragflow?color=blue&label=Rilis%20Terbaru" alt="Rilis Terbaru">
@@ -40,7 +43,7 @@
   <a href="https://github.com/infiniflow/ragflow/issues/12241">Peta Jalan</a> |
   <a href="https://twitter.com/infiniflowai">Twitter</a> |
   <a href="https://discord.gg/NjYzJD3GM3">Discord</a> |
-  <a href="https://demo.ragflow.io">Demo</a>
+  <a href="https://cloud.ragflow.io">Demo</a>
 </h4>
 
 <div align="center" style="margin-top:20px;margin-bottom:20px;">
@@ -76,7 +79,7 @@
 
 ## 🎮 Demo
 
-Coba demo kami di [https://demo.ragflow.io](https://demo.ragflow.io).
+Coba demo kami di [https://cloud.ragflow.io](https://cloud.ragflow.io).
 
 <div align="center" style="margin-top:20px;margin-bottom:20px;">
 <img src="https://raw.githubusercontent.com/infiniflow/ragflow-docs/refs/heads/image/image/chunking.gif" width="1200"/>
@@ -85,6 +88,7 @@ Coba demo kami di [https://demo.ragflow.io](https://demo.ragflow.io).
 
 ## 🔥 Pembaruan Terbaru
 
+- 2026-03-24 [RAGFlow Skill on OpenClaw](https://clawhub.ai/yingfeng/ragflow-skill) — Menyediakan skill resmi untuk mengakses dataset RAGFlow melalui OpenClaw.
 - 2025-12-26 Mendukung 'Memori' untuk agen AI.
 - 2025-11-19 Mendukung Gemini 3 Pro.
 - 2025-11-12 Mendukung sinkronisasi data dari Confluence, S3, Notion, Discord, Google Drive.
@@ -188,12 +192,12 @@ Coba demo kami di [https://demo.ragflow.io](https://demo.ragflow.io).
 > Semua gambar Docker dibangun untuk platform x86. Saat ini, kami tidak menawarkan gambar Docker untuk ARM64.
 > Jika Anda menggunakan platform ARM64, [silakan gunakan panduan ini untuk membangun gambar Docker yang kompatibel dengan sistem Anda](https://ragflow.io/docs/dev/build_docker_image).
 
-> Perintah di bawah ini mengunduh edisi v0.24.0 dari gambar Docker RAGFlow. Silakan merujuk ke tabel berikut untuk deskripsi berbagai edisi RAGFlow. Untuk mengunduh edisi RAGFlow yang berbeda dari v0.24.0, perbarui variabel RAGFLOW_IMAGE di docker/.env sebelum menggunakan docker compose untuk memulai server.
+> Perintah di bawah ini mengunduh edisi v0.25.0 dari gambar Docker RAGFlow. Silakan merujuk ke tabel berikut untuk deskripsi berbagai edisi RAGFlow. Untuk mengunduh edisi RAGFlow yang berbeda dari v0.25.0, perbarui variabel RAGFLOW_IMAGE di docker/.env sebelum menggunakan docker compose untuk memulai server.
 
 ```bash
    $ cd ragflow/docker
 
-   # git checkout v0.24.0
+   # git checkout v0.25.0
    # Opsional: gunakan tag stabil (lihat releases: https://github.com/infiniflow/ragflow/releases)
    # This steps ensures the **entrypoint.sh** file in the code matches the Docker image version.
 
@@ -299,7 +303,7 @@ docker build --platform linux/amd64 \
    git clone https://github.com/infiniflow/ragflow.git
    cd ragflow/
    uv sync --python 3.12 # install RAGFlow dependent python modules
-   uv run download_deps.py
+   uv run python3 download_deps.py
    pre-commit install
    ```
 3. Jalankan aplikasi yang diperlukan (MinIO, Elasticsearch, Redis, dan MySQL) menggunakan Docker Compose:
@@ -361,8 +365,8 @@ docker build --platform linux/amd64 \
 - [Quickstart](https://ragflow.io/docs/dev/)
 - [Configuration](https://ragflow.io/docs/dev/configurations)
 - [Release notes](https://ragflow.io/docs/dev/release_notes)
-- [User guides](https://ragflow.io/docs/dev/category/guides)
-- [Developer guides](https://ragflow.io/docs/dev/category/developers)
+- [User guides](https://ragflow.io/docs/category/user-guides)
+- [Developer guides](https://ragflow.io/docs/category/developer-guides)
 - [References](https://ragflow.io/docs/dev/category/references)
 - [FAQs](https://ragflow.io/docs/dev/faq)
 

README_ja.md

@@ -1,5 +1,5 @@
 <div align="center">
-<a href="https://demo.ragflow.io/">
+<a href="https://cloud.ragflow.io/">
 <img src="web/src/assets/logo-with-text.svg" width="350" alt="ragflow logo">
 </a>
 </div>
@@ -12,17 +12,20 @@
   <a href="./README_ko.md"><img alt="한국어" src="https://img.shields.io/badge/한국어-DFE0E5"></a>
   <a href="./README_id.md"><img alt="Bahasa Indonesia" src="https://img.shields.io/badge/Bahasa Indonesia-DFE0E5"></a>
   <a href="./README_pt_br.md"><img alt="Português(Brasil)" src="https://img.shields.io/badge/Português(Brasil)-DFE0E5"></a>
+  <a href="./README_fr.md"><img alt="README en Français" src="https://img.shields.io/badge/Français-DFE0E5"></a>
+  <a href="./README_ar.md"><img alt="README in Arabic" src="https://img.shields.io/badge/Arabic-DFE0E5"></a>
+  <a href="./README_tr.md"><img alt="Türkçe README" src="https://img.shields.io/badge/Türkçe-DFE0E5"></a>
 </p>
 
 <p align="center">
     <a href="https://x.com/intent/follow?screen_name=infiniflowai" target="_blank">
         <img src="https://img.shields.io/twitter/follow/infiniflow?logo=X&color=%20%23f5f5f5" alt="follow on X(Twitter)">
     </a>
-    <a href="https://demo.ragflow.io" target="_blank">
+    <a href="https://cloud.ragflow.io" target="_blank">
         <img alt="Static Badge" src="https://img.shields.io/badge/Online-Demo-4e6b99">
     </a>
     <a href="https://hub.docker.com/r/infiniflow/ragflow" target="_blank">
-        <img src="https://img.shields.io/docker/pulls/infiniflow/ragflow?label=Docker%20Pulls&color=0db7ed&logo=docker&logoColor=white&style=flat-square" alt="docker pull infiniflow/ragflow:v0.24.0">
+        <img src="https://img.shields.io/docker/pulls/infiniflow/ragflow?label=Docker%20Pulls&color=0db7ed&logo=docker&logoColor=white&style=flat-square" alt="docker pull infiniflow/ragflow:v0.25.0">
     </a>
     <a href="https://github.com/infiniflow/ragflow/releases/latest">
         <img src="https://img.shields.io/github/v/release/infiniflow/ragflow?color=blue&label=Latest%20Release" alt="Latest Release">
@@ -40,7 +43,7 @@
   <a href="https://github.com/infiniflow/ragflow/issues/12241">Roadmap</a> |
   <a href="https://twitter.com/infiniflowai">Twitter</a> |
   <a href="https://discord.gg/NjYzJD3GM3">Discord</a> |
-  <a href="https://demo.ragflow.io">Demo</a>
+  <a href="https://cloud.ragflow.io">Demo</a>
 </h4>
 
 <div align="center" style="margin-top:20px;margin-bottom:20px;">
@@ -57,7 +60,7 @@
 
 ## 🎮 Demo
 
-デモをお試しください：[https://demo.ragflow.io](https://demo.ragflow.io)。
+デモをお試しください：[https://cloud.ragflow.io](https://cloud.ragflow.io)。
 
 <div align="center" style="margin-top:20px;margin-bottom:20px;">
 <img src="https://raw.githubusercontent.com/infiniflow/ragflow-docs/refs/heads/image/image/chunking.gif" width="1200"/>
@@ -66,6 +69,7 @@
 
 ## 🔥 最新情報
 
+- 2026-03-24 [RAGFlow Skill on OpenClaw](https://clawhub.ai/yingfeng/ragflow-skill) — OpenClaw経由でRAGFlowデータセットにアクセスする公式スキルを提供。
 - 2025-12-26 AIエージェントの「メモリ」機能をサポート。
 - 2025-11-19 Gemini 3 Proをサポートしています。
 - 2025-11-12 Confluence、S3、Notion、Discord、Google Drive からのデータ同期をサポートします。
@@ -168,12 +172,12 @@
 > 現在、公式に提供されているすべての Docker イメージは x86 アーキテクチャ向けにビルドされており、ARM64 用の Docker イメージは提供されていません。
 > ARM64 アーキテクチャのオペレーティングシステムを使用している場合は、[このドキュメント](https://ragflow.io/docs/dev/build_docker_image)を参照して Docker イメージを自分でビルドしてください。
 
-> 以下のコマンドは、RAGFlow Docker イメージの v0.24.0 エディションをダウンロードします。異なる RAGFlow エディションの説明については、以下の表を参照してください。v0.24.0 とは異なるエディションをダウンロードするには、docker/.env ファイルの RAGFLOW_IMAGE 変数を適宜更新し、docker compose を使用してサーバーを起動してください。
+> 以下のコマンドは、RAGFlow Docker イメージの v0.25.0 エディションをダウンロードします。異なる RAGFlow エディションの説明については、以下の表を参照してください。v0.25.0 とは異なるエディションをダウンロードするには、docker/.env ファイルの RAGFLOW_IMAGE 変数を適宜更新し、docker compose を使用してサーバーを起動してください。
 
 ```bash
    $ cd ragflow/docker
 
-   # git checkout v0.24.0
+   # git checkout v0.25.0
    # 任意: 安定版タグを利用 (一覧: https://github.com/infiniflow/ragflow/releases)
    # この手順は、コード内の entrypoint.sh ファイルが Docker イメージのバージョンと一致していることを確認します。
 
@@ -299,7 +303,7 @@ docker build --platform linux/amd64 \
    git clone https://github.com/infiniflow/ragflow.git
    cd ragflow/
    uv sync --python 3.12 # install RAGFlow dependent python modules
-   uv run download_deps.py
+   uv run python3 download_deps.py
    pre-commit install
    ```
 3. Docker Compose を使用して依存サービス（MinIO、Elasticsearch、Redis、MySQL）を起動する:
@@ -361,8 +365,8 @@ docker build --platform linux/amd64 \
 - [Quickstart](https://ragflow.io/docs/dev/)
 - [Configuration](https://ragflow.io/docs/dev/configurations)
 - [Release notes](https://ragflow.io/docs/dev/release_notes)
-- [User guides](https://ragflow.io/docs/dev/category/guides)
-- [Developer guides](https://ragflow.io/docs/dev/category/developers)
+- [User guides](https://ragflow.io/docs/category/user-guides)
+- [Developer guides](https://ragflow.io/docs/category/developer-guides)
 - [References](https://ragflow.io/docs/dev/category/references)
 - [FAQs](https://ragflow.io/docs/dev/faq)
 

README_ko.md

@@ -1,5 +1,5 @@
 <div align="center">
-<a href="https://demo.ragflow.io/">
+<a href="https://cloud.ragflow.io/">
 <img src="web/src/assets/logo-with-text.svg" width="520" alt="ragflow logo">
 </a>
 </div>
@@ -12,17 +12,20 @@
   <a href="./README_ko.md"><img alt="한국어" src="https://img.shields.io/badge/한국어-DBEDFA"></a>
   <a href="./README_id.md"><img alt="Bahasa Indonesia" src="https://img.shields.io/badge/Bahasa Indonesia-DFE0E5"></a>
   <a href="./README_pt_br.md"><img alt="Português(Brasil)" src="https://img.shields.io/badge/Português(Brasil)-DFE0E5"></a>
+  <a href="./README_fr.md"><img alt="README en Français" src="https://img.shields.io/badge/Français-DFE0E5"></a>
+  <a href="./README_ar.md"><img alt="README in Arabic" src="https://img.shields.io/badge/Arabic-DFE0E5"></a>
+  <a href="./README_tr.md"><img alt="Türkçe README" src="https://img.shields.io/badge/Türkçe-DFE0E5"></a>
 </p>
 
 <p align="center">
     <a href="https://x.com/intent/follow?screen_name=infiniflowai" target="_blank">
         <img src="https://img.shields.io/twitter/follow/infiniflow?logo=X&color=%20%23f5f5f5" alt="follow on X(Twitter)">
     </a>
-    <a href="https://demo.ragflow.io" target="_blank">
+    <a href="https://cloud.ragflow.io" target="_blank">
         <img alt="Static Badge" src="https://img.shields.io/badge/Online-Demo-4e6b99">
     </a>
     <a href="https://hub.docker.com/r/infiniflow/ragflow" target="_blank">
-        <img src="https://img.shields.io/docker/pulls/infiniflow/ragflow?label=Docker%20Pulls&color=0db7ed&logo=docker&logoColor=white&style=flat-square" alt="docker pull infiniflow/ragflow:v0.24.0">
+        <img src="https://img.shields.io/docker/pulls/infiniflow/ragflow?label=Docker%20Pulls&color=0db7ed&logo=docker&logoColor=white&style=flat-square" alt="docker pull infiniflow/ragflow:v0.25.0">
     </a>
     <a href="https://github.com/infiniflow/ragflow/releases/latest">
         <img src="https://img.shields.io/github/v/release/infiniflow/ragflow?color=blue&label=Latest%20Release" alt="Latest Release">
@@ -40,7 +43,7 @@
   <a href="https://github.com/infiniflow/ragflow/issues/12241">Roadmap</a> |
   <a href="https://twitter.com/infiniflowai">Twitter</a> |
   <a href="https://discord.gg/NjYzJD3GM3">Discord</a> |
-  <a href="https://demo.ragflow.io">Demo</a>
+  <a href="https://cloud.ragflow.io">Demo</a>
 </h4>
 
 <div align="center" style="margin-top:20px;margin-bottom:20px;">
@@ -58,7 +61,7 @@
 
 ## 🎮 데모
 
-데모를 [https://demo.ragflow.io](https://demo.ragflow.io)에서 실행해 보세요.
+데모를 [https://cloud.ragflow.io](https://cloud.ragflow.io)에서 실행해 보세요.
 
 <div align="center" style="margin-top:20px;margin-bottom:20px;">
 <img src="https://raw.githubusercontent.com/infiniflow/ragflow-docs/refs/heads/image/image/chunking.gif" width="1200"/>
@@ -67,6 +70,7 @@
 
 ## 🔥 업데이트
 
+- 2026-03-24 [RAGFlow Skill on OpenClaw](https://clawhub.ai/yingfeng/ragflow-skill) — OpenClaw를 통해 RAGFlow 데이터셋에 접근하는 공식 스킬 제공.
 - 2025-12-26 AI 에이전트의 '메모리' 기능 지원.
 - 2025-11-19 Gemini 3 Pro를 지원합니다.
 - 2025-11-12 Confluence, S3, Notion, Discord, Google Drive에서 데이터 동기화를 지원합니다.
@@ -170,12 +174,12 @@
 > 모든 Docker 이미지는 x86 플랫폼을 위해 빌드되었습니다. 우리는 현재 ARM64 플랫폼을 위한 Docker 이미지를 제공하지 않습니다.
 > ARM64 플랫폼을 사용 중이라면, [시스템과 호환되는 Docker 이미지를 빌드하려면 이 가이드를 사용해 주세요](https://ragflow.io/docs/dev/build_docker_image).
 
-   > 아래 명령어는 RAGFlow Docker 이미지의 v0.24.0 버전을 다운로드합니다. 다양한 RAGFlow 버전에 대한 설명은 다음 표를 참조하십시오. v0.24.0과 다른 RAGFlow 버전을 다운로드하려면, docker/.env 파일에서 RAGFLOW_IMAGE 변수를 적절히 업데이트한 후 docker compose를 사용하여 서버를 시작하십시오.
+   > 아래 명령어는 RAGFlow Docker 이미지의 v0.25.0 버전을 다운로드합니다. 다양한 RAGFlow 버전에 대한 설명은 다음 표를 참조하십시오. v0.25.0과 다른 RAGFlow 버전을 다운로드하려면, docker/.env 파일에서 RAGFLOW_IMAGE 변수를 적절히 업데이트한 후 docker compose를 사용하여 서버를 시작하십시오.
 
    ```bash
    $ cd ragflow/docker
 
-   # git checkout v0.24.0
+   # git checkout v0.25.0
    # Optional: use a stable tag (see releases: https://github.com/infiniflow/ragflow/releases)
    # 이 단계는 코드의 entrypoint.sh 파일이 Docker 이미지 버전과 일치하도록 보장합니다.
 
@@ -294,7 +298,7 @@ docker build --platform linux/amd64 \
    git clone https://github.com/infiniflow/ragflow.git
    cd ragflow/
    uv sync --python 3.12 # install RAGFlow dependent python modules
-   uv run download_deps.py
+   uv run python3 download_deps.py
    pre-commit install
    ```
 
@@ -365,8 +369,8 @@ docker build --platform linux/amd64 \
 - [Quickstart](https://ragflow.io/docs/dev/)
 - [Configuration](https://ragflow.io/docs/dev/configurations)
 - [Release notes](https://ragflow.io/docs/dev/release_notes)
-- [User guides](https://ragflow.io/docs/dev/category/guides)
-- [Developer guides](https://ragflow.io/docs/dev/category/developers)
+- [User guides](https://ragflow.io/docs/category/user-guides)
+- [Developer guides](https://ragflow.io/docs/category/developer-guides)
 - [References](https://ragflow.io/docs/dev/category/references)
 - [FAQs](https://ragflow.io/docs/dev/faq)
 

README_pt_br.md

@@ -1,5 +1,5 @@
 <div align="center">
-<a href="https://demo.ragflow.io/">
+<a href="https://cloud.ragflow.io/">
 <img src="web/src/assets/logo-with-text.svg" width="520" alt="ragflow logo">
 </a>
 </div>
@@ -12,17 +12,20 @@
   <a href="./README_ko.md"><img alt="한국어" src="https://img.shields.io/badge/한국어-DFE0E5"></a>
   <a href="./README_id.md"><img alt="Bahasa Indonesia" src="https://img.shields.io/badge/Bahasa Indonesia-DFE0E5"></a>
   <a href="./README_pt_br.md"><img alt="Português(Brasil)" src="https://img.shields.io/badge/Português(Brasil)-DBEDFA"></a>
+  <a href="./README_fr.md"><img alt="README en Français" src="https://img.shields.io/badge/Français-DFE0E5"></a>
+  <a href="./README_ar.md"><img alt="README in Arabic" src="https://img.shields.io/badge/Arabic-DFE0E5"></a>
+  <a href="./README_tr.md"><img alt="Türkçe README" src="https://img.shields.io/badge/Türkçe-DFE0E5"></a>
 </p>
 
 <p align="center">
     <a href="https://x.com/intent/follow?screen_name=infiniflowai" target="_blank">
         <img src="https://img.shields.io/twitter/follow/infiniflow?logo=X&color=%20%23f5f5f5" alt="seguir no X(Twitter)">
     </a>
-    <a href="https://demo.ragflow.io" target="_blank">
+    <a href="https://cloud.ragflow.io" target="_blank">
         <img alt="Badge Estático" src="https://img.shields.io/badge/Online-Demo-4e6b99">
     </a>
     <a href="https://hub.docker.com/r/infiniflow/ragflow" target="_blank">
-        <img src="https://img.shields.io/docker/pulls/infiniflow/ragflow?label=Docker%20Pulls&color=0db7ed&logo=docker&logoColor=white&style=flat-square" alt="docker pull infiniflow/ragflow:v0.24.0">
+        <img src="https://img.shields.io/docker/pulls/infiniflow/ragflow?label=Docker%20Pulls&color=0db7ed&logo=docker&logoColor=white&style=flat-square" alt="docker pull infiniflow/ragflow:v0.25.0">
     </a>
     <a href="https://github.com/infiniflow/ragflow/releases/latest">
         <img src="https://img.shields.io/github/v/release/infiniflow/ragflow?color=blue&label=Última%20Relese" alt="Última Versão">
@@ -40,7 +43,7 @@
   <a href="https://github.com/infiniflow/ragflow/issues/12241">Roadmap</a> |
   <a href="https://twitter.com/infiniflowai">Twitter</a> |
   <a href="https://discord.gg/NjYzJD3GM3">Discord</a> |
-  <a href="https://demo.ragflow.io">Demo</a>
+  <a href="https://cloud.ragflow.io">Demo</a>
 </h4>
 
 <div align="center" style="margin-top:20px;margin-bottom:20px;">
@@ -77,7 +80,7 @@
 
 ## 🎮 Demo
 
-Experimente nossa demo em [https://demo.ragflow.io](https://demo.ragflow.io).
+Experimente nossa demo em [https://cloud.ragflow.io](https://cloud.ragflow.io).
 
 <div align="center" style="margin-top:20px;margin-bottom:20px;">
 <img src="https://raw.githubusercontent.com/infiniflow/ragflow-docs/refs/heads/image/image/chunking.gif" width="1200"/>
@@ -86,6 +89,7 @@ Experimente nossa demo em [https://demo.ragflow.io](https://demo.ragflow.io).
 
 ## 🔥 Últimas Atualizações
 
+- 24-03-2026 [RAGFlow Skill on OpenClaw](https://clawhub.ai/yingfeng/ragflow-skill) — Fornece um skill oficial para acessar datasets do RAGFlow via OpenClaw.
 - 26-12-2025 Suporte à função 'Memória' para agentes de IA.
 - 19-11-2025 Suporta Gemini 3 Pro.
 - 12-11-2025 Suporta a sincronização de dados do Confluence, S3, Notion, Discord e Google Drive.
@@ -188,12 +192,12 @@ Experimente nossa demo em [https://demo.ragflow.io](https://demo.ragflow.io).
 > Todas as imagens Docker são construídas para plataformas x86. Atualmente, não oferecemos imagens Docker para ARM64.
 > Se você estiver usando uma plataforma ARM64, por favor, utilize [este guia](https://ragflow.io/docs/dev/build_docker_image) para construir uma imagem Docker compatível com o seu sistema.
 
-    > O comando abaixo baixa a edição`v0.24.0` da imagem Docker do RAGFlow. Consulte a tabela a seguir para descrições de diferentes edições do RAGFlow. Para baixar uma edição do RAGFlow diferente da `v0.24.0`, atualize a variável `RAGFLOW_IMAGE` conforme necessário no **docker/.env** antes de usar `docker compose` para iniciar o servidor.
+    > O comando abaixo baixa a edição`v0.25.0` da imagem Docker do RAGFlow. Consulte a tabela a seguir para descrições de diferentes edições do RAGFlow. Para baixar uma edição do RAGFlow diferente da `v0.25.0`, atualize a variável `RAGFLOW_IMAGE` conforme necessário no **docker/.env** antes de usar `docker compose` para iniciar o servidor.
 
 ```bash
    $ cd ragflow/docker
 
-   # git checkout v0.24.0
+   # git checkout v0.25.0
    # Opcional: use uma tag estável (veja releases: https://github.com/infiniflow/ragflow/releases)
    # Esta etapa garante que o arquivo entrypoint.sh no código corresponda à versão da imagem do Docker.
 
@@ -316,7 +320,7 @@ docker build --platform linux/amd64 \
    git clone https://github.com/infiniflow/ragflow.git
    cd ragflow/
    uv sync --python 3.12 # instala os módulos Python dependentes do RAGFlow
-   uv run download_deps.py
+   uv run python3 download_deps.py
    pre-commit install
    ```
 3. Inicie os serviços dependentes (MinIO, Elasticsearch, Redis e MySQL) usando Docker Compose:
@@ -378,8 +382,8 @@ docker build --platform linux/amd64 \
 - [Quickstart](https://ragflow.io/docs/dev/)
 - [Configuration](https://ragflow.io/docs/dev/configurations)
 - [Release notes](https://ragflow.io/docs/dev/release_notes)
-- [User guides](https://ragflow.io/docs/dev/category/guides)
-- [Developer guides](https://ragflow.io/docs/dev/category/developers)
+- [User guides](https://ragflow.io/docs/category/user-guides)
+- [Developer guides](https://ragflow.io/docs/category/developer-guides)
 - [References](https://ragflow.io/docs/dev/category/references)
 - [FAQs](https://ragflow.io/docs/dev/faq)
 

README_tr.md

@@ -0,0 +1,409 @@
+<div align="center">
+<a href="https://cloud.ragflow.io/">
+<img src="web/src/assets/logo-with-text.svg" width="520" alt="ragflow logo">
+</a>
+</div>
+
+<p align="center">
+  <a href="./README.md"><img alt="README in English" src="https://img.shields.io/badge/English-DFE0E5"></a>
+  <a href="./README_zh.md"><img alt="简体中文版自述文件" src="https://img.shields.io/badge/简体中文-DFE0E5"></a>
+  <a href="./README_tzh.md"><img alt="繁體版中文自述文件" src="https://img.shields.io/badge/繁體中文-DFE0E5"></a>
+  <a href="./README_ja.md"><img alt="日本語のREADME" src="https://img.shields.io/badge/日本語-DFE0E5"></a>
+  <a href="./README_ko.md"><img alt="한국어" src="https://img.shields.io/badge/한국어-DFE0E5"></a>
+  <a href="./README_id.md"><img alt="Bahasa Indonesia" src="https://img.shields.io/badge/Bahasa Indonesia-DFE0E5"></a>
+  <a href="./README_pt_br.md"><img alt="Português(Brasil)" src="https://img.shields.io/badge/Português(Brasil)-DFE0E5"></a>
+  <a href="./README_fr.md"><img alt="README en Français" src="https://img.shields.io/badge/Français-DFE0E5"></a>
+  <a href="./README_ar.md"><img alt="README in Arabic" src="https://img.shields.io/badge/Arabic-DFE0E5"></a>
+  <a href="./README_tr.md"><img alt="Türkçe README" src="https://img.shields.io/badge/Türkçe-DBEDFA"></a>
+</p>
+
+<p align="center">
+    <a href="https://x.com/intent/follow?screen_name=infiniflowai" target="_blank">
+        <img src="https://img.shields.io/twitter/follow/infiniflow?logo=X&color=%20%23f5f5f5" alt="X(Twitter)'da takip et">
+    </a>
+    <a href="https://cloud.ragflow.io" target="_blank">
+        <img alt="Çevrimiçi Demo" src="https://img.shields.io/badge/Online-Demo-4e6b99">
+    </a>
+    <a href="https://hub.docker.com/r/infiniflow/ragflow" target="_blank">
+        <img src="https://img.shields.io/docker/pulls/infiniflow/ragflow?label=Docker%20Pulls&color=0db7ed&logo=docker&logoColor=white&style=flat-square" alt="docker pull infiniflow/ragflow:v0.25.0">
+    </a>
+    <a href="https://github.com/infiniflow/ragflow/releases/latest">
+        <img src="https://img.shields.io/github/v/release/infiniflow/ragflow?color=blue&label=Son%20Sürüm" alt="Son Sürüm">
+    </a>
+    <a href="https://github.com/infiniflow/ragflow/blob/main/LICENSE">
+        <img height="21" src="https://img.shields.io/badge/Lisans-Apache--2.0-ffffff?labelColor=d4eaf7&color=2e6cc4" alt="lisans">
+    </a>
+    <a href="https://deepwiki.com/infiniflow/ragflow">
+        <img alt="Ask DeepWiki" src="https://deepwiki.com/badge.svg">
+    </a>
+</p>
+
+<h4 align="center">
+  <a href="https://ragflow.io/docs/dev/">Dokümantasyon</a> |
+  <a href="https://github.com/infiniflow/ragflow/issues/12241">Yol Haritası</a> |
+  <a href="https://twitter.com/infiniflowai">Twitter</a> |
+  <a href="https://discord.gg/NjYzJD3GM3">Discord</a> |
+  <a href="https://cloud.ragflow.io">Demo</a>
+</h4>
+
+<div align="center" style="margin-top:20px;margin-bottom:20px;">
+<img src="https://raw.githubusercontent.com/infiniflow/ragflow-docs/refs/heads/image/image/ragflow-octoverse.png" width="1200"/>
+</div>
+
+<div align="center">
+<a href="https://trendshift.io/repositories/9064" target="_blank"><img src="https://trendshift.io/api/badge/repositories/9064" alt="infiniflow%2Fragflow | Trendshift" style="width: 250px; height: 55px;" width="250" height="55"/></a>
+</div>
+
+<details open>
+<summary><b>📕 İçindekiler</b></summary>
+
+- 💡 [RAGFlow Nedir?](#-ragflow-nedir)
+- 🎮 [Demo](#-demo)
+- 📌 [Son Güncellemeler](#-son-güncellemeler)
+- 🌟 [Temel Özellikler](#-temel-özellikler)
+- 🔎 [Sistem Mimarisi](#-sistem-mimarisi)
+- 🎬 [Başlarken](#-başlarken)
+- 🔧 [Yapılandırmalar](#-yapılandırmalar)
+- 🔧 [Docker İmajı Oluşturma](#-docker-i̇majı-oluşturma)
+- 🔨 [Geliştirme İçin Kaynaktan Hizmet Başlatma](#-geliştirme-i̇çin-kaynaktan-hizmet-başlatma)
+- 📚 [Dokümantasyon](#-dokümantasyon)
+- 📜 [Yol Haritası](#-yol-haritası)
+- 🏄 [Topluluk](#-topluluk)
+- 🙌 [Katkıda Bulunma](#-katkıda-bulunma)
+
+</details>
+
+## 💡 RAGFlow Nedir?
+
+[RAGFlow](https://ragflow.io/), derin doküman anlayışına dayalı, açık kaynaklı ve öncü bir Artırılmış Üretim ile Bilgi Erişimi ([RAG](https://ragflow.io/basics/what-is-rag)) motorudur. En son RAG teknolojisini Ajan yetenekleriyle birleştirerek LLM'ler için üstün bir bağlam katmanı oluşturur. Her ölçekteki kuruluşa uyarlanabilir, kolaylaştırılmış bir RAG iş akışı sunar. Yakınsanmış bir [bağlam motoru](https://ragflow.io/basics/what-is-agent-context-engine) ve hazır ajan şablonlarıyla donatılmış RAGFlow, geliştiricilerin karmaşık verileri yüksek doğrulukta, üretime hazır yapay zeka sistemlerine olağanüstü verimlilik ve hassasiyetle dönüştürmesini sağlar.
+
+## 🎮 Demo
+
+Demomuzu [https://cloud.ragflow.io](https://cloud.ragflow.io) adresinden deneyebilirsiniz.
+
+<div align="center" style="margin-top:20px;margin-bottom:20px;">
+<img src="https://raw.githubusercontent.com/infiniflow/ragflow-docs/refs/heads/image/image/chunking.gif" width="1200"/>
+<img src="https://raw.githubusercontent.com/infiniflow/ragflow-docs/refs/heads/image/image/agentic-dark.gif" width="1200"/>
+</div>
+
+## 🔥 Son Güncellemeler
+
+- 2026-03-24 [RAGFlow Skill on OpenClaw](https://clawhub.ai/yingfeng/ragflow-skill) — OpenClaw üzerinden RAGFlow veri setlerine erişmek için resmi bir skill sağlar.
+- 2025-12-26 Yapay zeka ajanı için 'Bellek' desteği eklendi.
+- 2025-11-19 Gemini 3 Pro desteği eklendi.
+- 2025-11-12 Confluence, S3, Notion, Discord, Google Drive'dan veri senkronizasyonu desteği eklendi.
+- 2025-10-23 Doküman ayrıştırma yöntemi olarak MinerU ve Docling desteği eklendi.
+- 2025-10-15 Düzenlenebilir veri alım hattı desteği eklendi.
+- 2025-08-08 OpenAI'ın en yeni GPT-5 serisi modelleri için destek eklendi.
+- 2025-08-01 Ajanlı iş akışı ve MCP desteği eklendi.
+- 2025-05-23 Ajana Python/JavaScript kod çalıştırıcı bileşeni eklendi.
+- 2025-05-05 Diller arası sorgu desteği eklendi.
+- 2025-03-19 PDF veya DOCX dosyalarındaki görselleri yorumlamak için çok modlu model desteği eklendi.
+
+## 🎉 Bizi Takip Edin
+
+⭐️ Heyecan verici yeni özellikler ve iyileştirmelerden haberdar olmak için depomuzı yıldızlayın! Yeni sürümler için anında bildirim alın! 🌟
+
+<div align="center" style="margin-top:20px;margin-bottom:20px;">
+<img src="https://github.com/user-attachments/assets/18c9707e-b8aa-4caf-a154-037089c105ba" width="1200"/>
+</div>
+
+## 🌟 Temel Özellikler
+
+### 🍭 **"Kaliteli girdi, kaliteli çıktı"**
+
+- Karmaşık formatlara sahip yapılandırılmamış verilerden [derin doküman anlayışı](./deepdoc/README.md) tabanlı bilgi çıkarımı.
+- Kelimenin tam anlamıyla sınırsız token içinde "samanlıkta iğne bulma" yeteneği.
+
+### 🍱 **Şablon tabanlı parçalama**
+
+- Akıllı ve açıklanabilir.
+- Aralarından seçim yapabileceğiniz çok sayıda şablon seçeneği.
+
+### 🌱 **Azaltılmış halüsinasyonlarla temellendirilmiş alıntılar**
+
+- İnsan müdahalesine olanak tanıyan metin parçalama görselleştirmesi.
+- Temellendirilmiş yanıtları desteklemek için anahtar referansların hızlı görüntülenmesi ve izlenebilir alıntılar.
+
+### 🍔 **Heterojen veri kaynaklarıyla uyumluluk**
+
+- Word, slaytlar, Excel, txt, görseller, taranmış kopyalar, yapılandırılmış veriler, web sayfaları ve daha fazlasını destekler.
+
+### 🛀 **Otomatik ve zahmetsiz RAG iş akışı**
+
+- Hem bireysel hem de büyük işletmeler için özelleştirilmiş kolaylaştırılmış RAG düzenlemesi.
+- Yapılandırılabilir LLM'ler ve gömme (embedding) modelleri.
+- Birleştirilmiş yeniden sıralama ile çoklu geri çağırma.
+- İş süreçlerine sorunsuz entegrasyon için sezgisel API'ler.
+
+## 🔎 Sistem Mimarisi
+
+<div align="center" style="margin-top:20px;margin-bottom:20px;">
+<img src="https://github.com/user-attachments/assets/31b0dd6f-ca4f-445a-9457-70cb44a381b2" width="1000"/>
+</div>
+
+## 🎬 Başlarken
+
+### 📝 Ön Koşullar
+
+- CPU >= 4 çekirdek
+- RAM >= 16 GB
+- Disk >= 50 GB
+- Docker >= 24.0.0 & Docker Compose >= v2.26.1
+- [gVisor](https://gvisor.dev/docs/user_guide/install/): Yalnızca RAGFlow'un kod çalıştırıcı (sandbox) özelliğini kullanmayı planlıyorsanız gereklidir.
+
+> [!TIP]
+> Yerel makinenize (Windows, Mac veya Linux) Docker yüklemediyseniz, [Docker Engine Kurulumu](https://docs.docker.com/engine/install/) sayfasına bakın.
+
+### 🚀 Sunucuyu Başlatma
+
+1. `vm.max_map_count` değerinin >= 262144 olduğundan emin olun:
+
+   > `vm.max_map_count` değerini kontrol etmek için:
+   >
+   > ```bash
+   > $ sysctl vm.max_map_count
+   > ```
+   >
+   > Değer 262144'ten düşükse, en az 262144 olarak ayarlayın.
+   >
+   > ```bash
+   > # Bu örnekte 262144 olarak ayarlıyoruz:
+   > $ sudo sysctl -w vm.max_map_count=262144
+   > ```
+   >
+   > Bu değişiklik sistem yeniden başlatıldığında sıfırlanacaktır. Değişikliğin kalıcı olmasını sağlamak için
+   > **/etc/sysctl.conf** dosyasındaki `vm.max_map_count` değerini buna göre ekleyin veya güncelleyin:
+   >
+   > ```bash
+   > vm.max_map_count=262144
+   > ```
+   >
+2. Depoyu klonlayın:
+
+   ```bash
+   $ git clone https://github.com/infiniflow/ragflow.git
+   ```
+3. Önceden oluşturulmuş Docker imajlarını kullanarak sunucuyu başlatın:
+
+> [!CAUTION]
+> Tüm Docker imajları x86 platformları için oluşturulmuştur. Şu anda ARM64 için Docker imajı sunmuyoruz.
+> ARM64 platformundaysanız, sisteminizle uyumlu bir Docker imajı oluşturmak için [bu kılavuzu](https://ragflow.io/docs/dev/build_docker_image) takip edin.
+
+> Aşağıdaki komut RAGFlow Docker imajının `v0.25.0` sürümünü indirir. Farklı RAGFlow sürümleri için aşağıdaki tabloya bakın. `v0.25.0` dışında bir sürüm indirmek için, `docker compose` ile sunucuyu başlatmadan önce **docker/.env** dosyasındaki `RAGFLOW_IMAGE` değişkenini güncelleyin.
+
+```bash
+   $ cd ragflow/docker
+
+   # git checkout v0.25.0
+   # İsteğe bağlı: Kararlı bir etiket kullanın (sürümler: https://github.com/infiniflow/ragflow/releases)
+   # Bu adım, koddaki **entrypoint.sh** dosyasının Docker imaj sürümüyle eşleşmesini sağlar.
+
+   # DeepDoc görevleri için CPU kullanımı:
+   $ docker compose -f docker-compose.yml up -d
+
+   # DeepDoc görevlerini hızlandırmak için GPU kullanımı:
+   # sed -i '1i DEVICE=gpu' .env
+   # docker compose -f docker-compose.yml up -d
+```
+
+> Not: `v0.22.0` öncesinde hem gömme modelleri içeren imajlar hem de gömme modelleri içermeyen ince (slim) imajlar sunuyorduk. Detaylar aşağıdadır:
+
+| RAGFlow imaj etiketi | İmaj boyutu (GB) | Gömme modelleri var mı? | Kararlı mı?    |
+|-----------------------|-------------------|-------------------------|-----------------|
+| v0.21.1               | &approx;9        | ✔️                      | Kararlı sürüm   |
+| v0.21.1-slim          | &approx;2        | ❌                       | Kararlı sürüm   |
+
+> `v0.22.0`'dan itibaren yalnızca ince (slim) sürümü sunuyoruz ve imaj etiketine artık **-slim** son eki eklemiyoruz.
+
+4. Sunucu çalışır duruma geldikten sonra sunucu durumunu kontrol edin:
+
+   ```bash
+   $ docker logs -f docker-ragflow-cpu-1
+   ```
+
+   _Aşağıdaki çıktı, sistemin başarıyla başlatıldığını onaylar:_
+
+   ```bash
+
+         ____   ___    ______ ______ __
+        / __ \ /   |  / ____// ____// /____  _      __
+       / /_/ // /| | / / __ / /_   / // __ \| | /| / /
+      / _, _// ___ |/ /_/ // __/  / // /_/ /| |/ |/ /
+     /_/ |_|/_/  |_|\____//_/    /_/ \____/ |__/|__/
+
+    * Running on all addresses (0.0.0.0)
+   ```
+
+   > Bu onay adımını atlayıp doğrudan RAGFlow'a giriş yaparsanız, o anda RAGFlow tam olarak başlatılmamış olabileceğinden
+   > tarayıcınız `ağ hatası` uyarısı verebilir.
+   >
+5. Web tarayıcınıza sunucunuzun IP adresini girin ve RAGFlow'a giriş yapın.
+
+   > Varsayılan ayarlarla, yalnızca `http://MAKİNENİZİN_IP_ADRESİ` girmeniz yeterlidir (port numarası **gerekmez**),
+   > çünkü varsayılan HTTP sunucu portu `80` varsayılan yapılandırmalar kullanıldığında ihmal edilebilir.
+   >
+6. [service_conf.yaml.template](./docker/service_conf.yaml.template) dosyasında, `user_default_llm` içinde istediğiniz LLM sağlayıcısını seçin ve
+   `API_KEY` alanını ilgili API anahtarıyla güncelleyin.
+
+   > Daha fazla bilgi için [llm_api_key_setup](https://ragflow.io/docs/dev/llm_api_key_setup) sayfasına bakın.
+   >
+
+   _Gösteri başlasın!_
+
+## 🔧 Yapılandırmalar
+
+Sistem yapılandırmaları söz konusu olduğunda, aşağıdaki dosyaları yönetmeniz gerekecektir:
+
+- [.env](./docker/.env): `SVR_HTTP_PORT`, `MYSQL_PASSWORD` ve `MINIO_PASSWORD` gibi temel sistem ayarlarını içerir.
+- [service_conf.yaml.template](./docker/service_conf.yaml.template): Arka uç hizmetlerini yapılandırır. Bu dosyadaki ortam değişkenleri, Docker konteyneri başladığında otomatik olarak doldurulacaktır. Docker konteyneri içinde ayarlanan tüm ortam değişkenleri kullanıma hazır olacak ve hizmet davranışını dağıtım ortamına göre özelleştirmenize olanak tanıyacaktır.
+- [docker-compose.yml](./docker/docker-compose.yml): Sistem, başlatılmak için [docker-compose.yml](./docker/docker-compose.yml) dosyasına dayanır.
+
+> [./docker/README](./docker/README.md) dosyası, [service_conf.yaml.template](./docker/service_conf.yaml.template) dosyasında `${ENV_VARS}` olarak kullanılabilen ortam ayarları ve hizmet yapılandırmalarının ayrıntılı bir açıklamasını sağlar.
+
+Varsayılan HTTP sunucu portunu (80) değiştirmek için [docker-compose.yml](./docker/docker-compose.yml) dosyasında `80:80` ifadesini `<SUNUCU_PORTUNUZ>:80` olarak değiştirin.
+
+Yukarıdaki yapılandırma değişikliklerinin etkili olması için tüm konteynerlerin yeniden başlatılması gerekir:
+
+> ```bash
+> $ docker compose -f docker-compose.yml up -d
+> ```
+
+### Doküman Motorunu Elasticsearch'ten Infinity'ye Geçirme
+
+RAGFlow varsayılan olarak tam metin ve vektörlerin depolanması için Elasticsearch kullanır. [Infinity](https://github.com/infiniflow/infinity/)'ye geçmek için şu adımları izleyin:
+
+1. Çalışan tüm konteynerleri durdurun:
+
+   ```bash
+   $ docker compose -f docker/docker-compose.yml down -v
+   ```
+
+> [!WARNING]
+> `-v` seçeneği Docker konteyner birimlerini silecek ve mevcut veriler temizlenecektir.
+
+2. **docker/.env** dosyasında `DOC_ENGINE` değerini `infinity` olarak ayarlayın.
+3. Konteynerleri başlatın:
+
+   ```bash
+   $ docker compose -f docker-compose.yml up -d
+   ```
+
+> [!WARNING]
+> Linux/arm64 makinesinde Infinity'ye geçiş henüz resmi olarak desteklenmemektedir.
+
+## 🔧 Docker İmajı Oluşturma
+
+Bu imaj yaklaşık 2 GB boyutundadır ve harici LLM ile gömme hizmetlerine bağlıdır.
+
+```bash
+git clone https://github.com/infiniflow/ragflow.git
+cd ragflow/
+docker build --platform linux/amd64 -f Dockerfile -t infiniflow/ragflow:nightly .
+```
+
+Veya bir proxy arkasındaysanız, proxy parametrelerini iletebilirsiniz:
+
+```bash
+docker build --platform linux/amd64 \
+  --build-arg http_proxy=http://PROXY_ADRESINIZ:PORT \
+  --build-arg https_proxy=http://PROXY_ADRESINIZ:PORT \
+  -f Dockerfile -t infiniflow/ragflow:nightly .
+```
+
+## 🔨 Geliştirme İçin Kaynaktan Hizmet Başlatma
+
+1. `uv` ve `pre-commit` yükleyin veya zaten yüklüyse bu adımı atlayın:
+
+   ```bash
+   pipx install uv pre-commit
+   ```
+2. Kaynak kodunu klonlayın ve Python bağımlılıklarını yükleyin:
+
+   ```bash
+   git clone https://github.com/infiniflow/ragflow.git
+   cd ragflow/
+   uv sync --python 3.12 # RAGFlow'un bağımlı Python modüllerini yükler
+   uv run python3 download_deps.py
+   pre-commit install
+   ```
+3. Bağımlı hizmetleri (MinIO, Elasticsearch, Redis ve MySQL) Docker Compose kullanarak başlatın:
+
+   ```bash
+   docker compose -f docker/docker-compose-base.yml up -d
+   ```
+
+   **docker/.env** dosyasında belirtilen tüm ana bilgisayar adlarını `127.0.0.1`'e çözümlemek için `/etc/hosts` dosyasına aşağıdaki satırı ekleyin:
+
+   ```
+   127.0.0.1       es01 infinity mysql minio redis sandbox-executor-manager
+   ```
+4. HuggingFace'e erişemiyorsanız, bir ayna site kullanmak için `HF_ENDPOINT` ortam değişkenini ayarlayın:
+
+   ```bash
+   export HF_ENDPOINT=https://hf-mirror.com
+   ```
+5. İşletim sisteminizde jemalloc yoksa, aşağıdaki şekilde yükleyin:
+
+   ```bash
+   # Ubuntu
+   sudo apt-get install libjemalloc-dev
+   # CentOS
+   sudo yum install jemalloc
+   # OpenSUSE
+   sudo zypper install jemalloc
+   # macOS
+   sudo brew install jemalloc
+   ```
+6. Arka uç hizmetini başlatın:
+
+   ```bash
+   source .venv/bin/activate
+   export PYTHONPATH=$(pwd)
+   bash docker/launch_backend_service.sh
+   ```
+7. Ön yüz bağımlılıklarını yükleyin:
+
+   ```bash
+   cd web
+   npm install
+   ```
+8. Ön yüz hizmetini başlatın:
+
+   ```bash
+   npm run dev
+   ```
+
+   _Aşağıdaki çıktı, sistemin başarıyla başlatıldığını onaylar:_
+
+   ![](https://github.com/user-attachments/assets/0daf462c-a24d-4496-a66f-92533534e187)
+9. Geliştirme tamamlandıktan sonra RAGFlow ön yüz ve arka uç hizmetini durdurun:
+
+   ```bash
+   pkill -f "ragflow_server.py|task_executor.py"
+   ```
+
+## 📚 Dokümantasyon
+
+- [Hızlı Başlangıç](https://ragflow.io/docs/dev/)
+- [Yapılandırma](https://ragflow.io/docs/dev/configurations)
+- [Sürüm Notları](https://ragflow.io/docs/dev/release_notes)
+- [Kullanıcı Kılavuzları](https://ragflow.io/docs/category/user-guides)
+- [Geliştirici Kılavuzları](https://ragflow.io/docs/category/developer-guides)
+- [Referanslar](https://ragflow.io/docs/dev/category/references)
+- [SSS](https://ragflow.io/docs/dev/faq)
+
+## 📜 Yol Haritası
+
+[RAGFlow Yol Haritası 2026](https://github.com/infiniflow/ragflow/issues/12241) sayfasına bakın.
+
+## 🏄 Topluluk
+
+- [Discord](https://discord.gg/NjYzJD3GM3)
+- [Twitter](https://twitter.com/infiniflowai)
+- [GitHub Tartışmalar](https://github.com/orgs/infiniflow/discussions)
+
+## 🙌 Katkıda Bulunma
+
+RAGFlow, açık kaynak iş birliği sayesinde gelişmektedir. Bu anlayışla, topluluktan gelen çeşitli katkıları benimsiyoruz.
+Bir parçası olmak istiyorsanız, önce [Katkıda Bulunma Kılavuzumuzu](https://ragflow.io/docs/dev/contributing) inceleyin.

README_tzh.md

@@ -1,5 +1,5 @@
 <div align="center">
-<a href="https://demo.ragflow.io/">
+<a href="https://cloud.ragflow.io/">
 <img src="web/src/assets/logo-with-text.svg" width="350" alt="ragflow logo">
 </a>
 </div>
@@ -12,17 +12,20 @@
   <a href="./README_ko.md"><img alt="한국어" src="https://img.shields.io/badge/한국어-DFE0E5"></a>
   <a href="./README_id.md"><img alt="Bahasa Indonesia" src="https://img.shields.io/badge/Bahasa Indonesia-DFE0E5"></a>
   <a href="./README_pt_br.md"><img alt="Português(Brasil)" src="https://img.shields.io/badge/Português(Brasil)-DFE0E5"></a>
+  <a href="./README_fr.md"><img alt="README en Français" src="https://img.shields.io/badge/Français-DFE0E5"></a>
+  <a href="./README_ar.md"><img alt="README in Arabic" src="https://img.shields.io/badge/Arabic-DFE0E5"></a>
+  <a href="./README_tr.md"><img alt="Türkçe README" src="https://img.shields.io/badge/Türkçe-DFE0E5"></a>
 </p>
 
 <p align="center">
     <a href="https://x.com/intent/follow?screen_name=infiniflowai" target="_blank">
         <img src="https://img.shields.io/twitter/follow/infiniflow?logo=X&color=%20%23f5f5f5" alt="follow on X(Twitter)">
     </a>
-    <a href="https://demo.ragflow.io" target="_blank">
+    <a href="https://cloud.ragflow.io" target="_blank">
         <img alt="Static Badge" src="https://img.shields.io/badge/Online-Demo-4e6b99">
     </a>
     <a href="https://hub.docker.com/r/infiniflow/ragflow" target="_blank">
-        <img src="https://img.shields.io/docker/pulls/infiniflow/ragflow?label=Docker%20Pulls&color=0db7ed&logo=docker&logoColor=white&style=flat-square" alt="docker pull infiniflow/ragflow:v0.24.0">
+        <img src="https://img.shields.io/docker/pulls/infiniflow/ragflow?label=Docker%20Pulls&color=0db7ed&logo=docker&logoColor=white&style=flat-square" alt="docker pull infiniflow/ragflow:v0.25.0">
     </a>
     <a href="https://github.com/infiniflow/ragflow/releases/latest">
         <img src="https://img.shields.io/github/v/release/infiniflow/ragflow?color=blue&label=Latest%20Release" alt="Latest Release">
@@ -40,7 +43,7 @@
   <a href="https://github.com/infiniflow/ragflow/issues/12241">Roadmap</a> |
   <a href="https://twitter.com/infiniflowai">Twitter</a> |
   <a href="https://discord.gg/NjYzJD3GM3">Discord</a> |
-  <a href="https://demo.ragflow.io">Demo</a>
+  <a href="https://cloud.ragflow.io">Demo</a>
 </h4>
 
 <div align="center" style="margin-top:20px;margin-bottom:20px;">
@@ -76,7 +79,7 @@
 
 ## 🎮 Demo 試用
 
-請登入網址 [https://demo.ragflow.io](https://demo.ragflow.io) 試用 demo。
+請登入網址 [https://cloud.ragflow.io](https://cloud.ragflow.io) 試用 demo。
 
 <div align="center" style="margin-top:20px;margin-bottom:20px;">
 <img src="https://raw.githubusercontent.com/infiniflow/ragflow-docs/refs/heads/image/image/chunking.gif" width="1200"/>
@@ -85,6 +88,7 @@
 
 ## 🔥 近期更新
 
+- 2026-03-24 發布 [RAGFlow 官方 Skill](https://clawhub.ai/yingfeng/ragflow-skill) — 提供官方 Skill 以透過 OpenClaw 訪問 RAGFlow 數據集。
 - 2025-12-26 支援AI代理的「記憶」功能。
 - 2025-11-19 支援 Gemini 3 Pro。
 - 2025-11-12 支援從 Confluence、S3、Notion、Discord、Google Drive 進行資料同步。
@@ -187,12 +191,12 @@
 > 所有 Docker 映像檔都是為 x86 平台建置的。目前，我們不提供 ARM64 平台的 Docker 映像檔。
 > 如果您使用的是 ARM64 平台，請使用 [這份指南](https://ragflow.io/docs/dev/build_docker_image) 來建置適合您系統的 Docker 映像檔。
 
-> 執行以下指令會自動下載 RAGFlow Docker 映像 `v0.24.0`。請參考下表查看不同 Docker 發行版的說明。如需下載不同於 `v0.24.0` 的 Docker 映像，請在執行 `docker compose` 啟動服務之前先更新 **docker/.env** 檔案內的 `RAGFLOW_IMAGE` 變數。
+> 執行以下指令會自動下載 RAGFlow Docker 映像 `v0.25.0`。請參考下表查看不同 Docker 發行版的說明。如需下載不同於 `v0.25.0` 的 Docker 映像，請在執行 `docker compose` 啟動服務之前先更新 **docker/.env** 檔案內的 `RAGFLOW_IMAGE` 變數。
 
 ```bash
    $ cd ragflow/docker
 
-   # git checkout v0.24.0
+   # git checkout v0.25.0
    # 可選：使用穩定版標籤（查看發佈：https://github.com/infiniflow/ragflow/releases）
    # 此步驟確保程式碼中的 entrypoint.sh 檔案與 Docker 映像版本一致。
 
@@ -326,7 +330,7 @@ docker build --platform linux/amd64 \
    git clone https://github.com/infiniflow/ragflow.git
    cd ragflow/
    uv sync --python 3.12 # install RAGFlow dependent python modules
-   uv run download_deps.py
+   uv run python3 download_deps.py
    pre-commit install
    ```
 3. 透過 Docker Compose 啟動依賴的服務（MinIO, Elasticsearch, Redis, and MySQL）：
@@ -392,8 +396,8 @@ docker build --platform linux/amd64 \
 - [Quickstart](https://ragflow.io/docs/dev/)
 - [Configuration](https://ragflow.io/docs/dev/configurations)
 - [Release notes](https://ragflow.io/docs/dev/release_notes)
-- [User guides](https://ragflow.io/docs/dev/category/guides)
-- [Developer guides](https://ragflow.io/docs/dev/category/developers)
+- [User guides](https://ragflow.io/docs/category/user-guides)
+- [Developer guides](https://ragflow.io/docs/category/developer-guides)
 - [References](https://ragflow.io/docs/dev/category/references)
 - [FAQs](https://ragflow.io/docs/dev/faq)
 

README_zh.md

@@ -1,5 +1,5 @@
 <div align="center">
-<a href="https://demo.ragflow.io/">
+<a href="https://cloud.ragflow.io/">
 <img src="web/src/assets/logo-with-text.svg" width="350" alt="ragflow logo">
 </a>
 </div>
@@ -12,17 +12,20 @@
   <a href="./README_ko.md"><img alt="한국어" src="https://img.shields.io/badge/한국어-DFE0E5"></a>
   <a href="./README_id.md"><img alt="Bahasa Indonesia" src="https://img.shields.io/badge/Bahasa Indonesia-DFE0E5"></a>
   <a href="./README_pt_br.md"><img alt="Português(Brasil)" src="https://img.shields.io/badge/Português(Brasil)-DFE0E5"></a>
+  <a href="./README_fr.md"><img alt="README en Français" src="https://img.shields.io/badge/Français-DFE0E5"></a>
+  <a href="./README_ar.md"><img alt="README in Arabic" src="https://img.shields.io/badge/Arabic-DFE0E5"></a>
+  <a href="./README_tr.md"><img alt="Türkçe README" src="https://img.shields.io/badge/Türkçe-DFE0E5"></a>
 </p>
 
 <p align="center">
     <a href="https://x.com/intent/follow?screen_name=infiniflowai" target="_blank">
         <img src="https://img.shields.io/twitter/follow/infiniflow?logo=X&color=%20%23f5f5f5" alt="follow on X(Twitter)">
     </a>
-    <a href="https://demo.ragflow.io" target="_blank">
+    <a href="https://cloud.ragflow.io" target="_blank">
         <img alt="Static Badge" src="https://img.shields.io/badge/Online-Demo-4e6b99">
     </a>
     <a href="https://hub.docker.com/r/infiniflow/ragflow" target="_blank">
-        <img src="https://img.shields.io/docker/pulls/infiniflow/ragflow?label=Docker%20Pulls&color=0db7ed&logo=docker&logoColor=white&style=flat-square" alt="docker pull infiniflow/ragflow:v0.24.0">
+        <img src="https://img.shields.io/docker/pulls/infiniflow/ragflow?label=Docker%20Pulls&color=0db7ed&logo=docker&logoColor=white&style=flat-square" alt="docker pull infiniflow/ragflow:v0.25.0">
     </a>
     <a href="https://github.com/infiniflow/ragflow/releases/latest">
         <img src="https://img.shields.io/github/v/release/infiniflow/ragflow?color=blue&label=Latest%20Release" alt="Latest Release">
@@ -40,7 +43,7 @@
   <a href="https://github.com/infiniflow/ragflow/issues/12241">Roadmap</a> |
   <a href="https://twitter.com/infiniflowai">Twitter</a> |
   <a href="https://discord.gg/NjYzJD3GM3">Discord</a> |
-  <a href="https://demo.ragflow.io">Demo</a>
+  <a href="https://cloud.ragflow.io">Demo</a>
 </h4>
 
 <div align="center" style="margin-top:20px;margin-bottom:20px;">
@@ -76,7 +79,7 @@
 
 ## 🎮 Demo 试用
 
-请登录网址 [https://demo.ragflow.io](https://demo.ragflow.io) 试用 demo。
+请登录网址 [https://cloud.ragflow.io](https://cloud.ragflow.io) 试用 demo。
 
 <div align="center" style="margin-top:20px;margin-bottom:20px;">
 <img src="https://raw.githubusercontent.com/infiniflow/ragflow-docs/refs/heads/image/image/chunking.gif" width="1200"/>
@@ -85,7 +88,8 @@
 
 ## 🔥 近期更新
 
-- 2025-12-26 支持AI代理的“记忆”功能。
+- 2026-03-24 发布 [RAGFlow 官方 Skill](https://clawhub.ai/yingfeng/ragflow-skill) — 提供官方 Skill 以通过 OpenClaw 访问 RAGFlow 数据集。
+- 2025-12-26 支持AI代理的"记忆"功能。
 - 2025-11-19 支持 Gemini 3 Pro。
 - 2025-11-12 支持从 Confluence、S3、Notion、Discord、Google Drive 进行数据同步。
 - 2025-10-23 支持 MinerU 和 Docling 作为文档解析方法。
@@ -188,12 +192,12 @@
 > 请注意，目前官方提供的所有 Docker 镜像均基于 x86 架构构建，并不提供基于 ARM64 的 Docker 镜像。
 > 如果你的操作系统是 ARM64 架构，请参考[这篇文档](https://ragflow.io/docs/dev/build_docker_image)自行构建 Docker 镜像。
 
-   > 运行以下命令会自动下载 RAGFlow Docker 镜像 `v0.24.0`。请参考下表查看不同 Docker 发行版的描述。如需下载不同于 `v0.24.0` 的 Docker 镜像，请在运行 `docker compose` 启动服务之前先更新 **docker/.env** 文件内的 `RAGFLOW_IMAGE` 变量。
+   > 运行以下命令会自动下载 RAGFlow Docker 镜像 `v0.25.0`。请参考下表查看不同 Docker 发行版的描述。如需下载不同于 `v0.25.0` 的 Docker 镜像，请在运行 `docker compose` 启动服务之前先更新 **docker/.env** 文件内的 `RAGFLOW_IMAGE` 变量。
 
    ```bash
    $ cd ragflow/docker
 
-   # git checkout v0.24.0
+   # git checkout v0.25.0
    # 可选：使用稳定版本标签（查看发布：https://github.com/infiniflow/ragflow/releases）
    # 这一步确保代码中的 entrypoint.sh 文件与 Docker 镜像的版本保持一致。
 
@@ -326,7 +330,7 @@ docker build --platform linux/amd64 \
    git clone https://github.com/infiniflow/ragflow.git
    cd ragflow/
    uv sync --python 3.12 # install RAGFlow dependent python modules
-   uv run download_deps.py
+   uv run python3 download_deps.py
    pre-commit install
    ```
 
@@ -395,8 +399,8 @@ docker build --platform linux/amd64 \
 - [Quickstart](https://ragflow.io/docs/dev/)
 - [Configuration](https://ragflow.io/docs/dev/configurations)
 - [Release notes](https://ragflow.io/docs/dev/release_notes)
-- [User guides](https://ragflow.io/docs/dev/category/guides)
-- [Developer guides](https://ragflow.io/docs/dev/category/developers)
+- [User guides](https://ragflow.io/docs/category/user-guides)
+- [Developer guides](https://ragflow.io/docs/category/developer-guides)
 - [References](https://ragflow.io/docs/dev/category/references)
 - [FAQs](https://ragflow.io/docs/dev/faq)
 

admin/client/COMMAND.md

@@ -0,0 +1,779 @@
+# RAGFlow CLI User Command Reference
+
+This document describes the user commands available in RAGFlow CLI. All commands must end with a semicolon (`;`).
+
+## Command List
+
+### ping_server
+
+**Description**  
+Tests the connection status to the server.
+
+**Usage**  
+```
+PING;
+```
+
+**Parameters**  
+No parameters.
+
+**Example**  
+```
+ragflow> PING;
+```
+
+**Display Effect**  
+(Sample output will be provided by the user)
+
+---
+
+### show_current_user
+
+**Description**  
+Displays information about the currently logged-in user.
+
+**Usage**  
+```
+SHOW CURRENT USER;
+```
+
+**Parameters**  
+No parameters.
+
+**Example**  
+```
+ragflow> SHOW CURRENT USER;
+```
+
+**Display Effect**  
+(Sample output will be provided by the user)
+
+---
+
+### create_model_provider
+
+**Description**  
+Creates a new model provider.
+
+**Usage**  
+```
+CREATE MODEL PROVIDER <provider_name> <provider_key>;
+```
+
+**Parameters**  
+- `provider_name`: Provider name, quoted string.
+- `provider_key`: Provider key, quoted string.
+
+**Example**  
+```
+ragflow> CREATE MODEL PROVIDER 'openai' 'sk-...';
+```
+
+**Display Effect**  
+(Sample output will be provided by the user)
+
+---
+
+### drop_model_provider
+
+**Description**  
+Deletes a model provider.
+
+**Usage**  
+```
+DROP MODEL PROVIDER <provider_name>;
+```
+
+**Parameters**  
+- `provider_name`: Name of the provider to delete, quoted string.
+
+**Example**  
+```
+ragflow> DROP MODEL PROVIDER 'openai';
+```
+
+**Display Effect**  
+(Sample output will be provided by the user)
+
+---
+
+### set_default_llm
+
+**Description**  
+Sets the default LLM (Large Language Model).
+
+**Usage**  
+```
+SET DEFAULT LLM <llm_id>;
+```
+
+**Parameters**  
+- `llm_id`: LLM identifier, quoted string.
+
+**Example**  
+```
+ragflow> SET DEFAULT LLM 'gpt-4';
+```
+
+**Display Effect**  
+(Sample output will be provided by the user)
+
+---
+
+### set_default_vlm
+
+**Description**  
+Sets the default VLM (Vision Language Model).
+
+**Usage**  
+```
+SET DEFAULT VLM <vlm_id>;
+```
+
+**Parameters**  
+- `vlm_id`: VLM identifier, quoted string.
+
+**Example**  
+```
+ragflow> SET DEFAULT VLM 'clip-vit-large';
+```
+
+**Display Effect**  
+(Sample output will be provided by the user)
+
+---
+
+### set_default_embedding
+
+**Description**  
+Sets the default embedding model.
+
+**Usage**  
+```
+SET DEFAULT EMBEDDING <embedding_id>;
+```
+
+**Parameters**  
+- `embedding_id`: Embedding model identifier, quoted string.
+
+**Example**  
+```
+ragflow> SET DEFAULT EMBEDDING 'text-embedding-ada-002';
+```
+
+**Display Effect**  
+(Sample output will be provided by the user)
+
+---
+
+### set_default_reranker
+
+**Description**  
+Sets the default reranker model.
+
+**Usage**  
+```
+SET DEFAULT RERANKER <reranker_id>;
+```
+
+**Parameters**  
+- `reranker_id`: Reranker model identifier, quoted string.
+
+**Example**  
+```
+ragflow> SET DEFAULT RERANKER 'bge-reranker-large';
+```
+
+**Display Effect**  
+(Sample output will be provided by the user)
+
+---
+
+### set_default_asr
+
+**Description**  
+Sets the default ASR (Automatic Speech Recognition) model.
+
+**Usage**  
+```
+SET DEFAULT ASR <asr_id>;
+```
+
+**Parameters**  
+- `asr_id`: ASR model identifier, quoted string.
+
+**Example**  
+```
+ragflow> SET DEFAULT ASR 'whisper-large';
+```
+
+**Display Effect**  
+(Sample output will be provided by the user)
+
+---
+
+### set_default_tts
+
+**Description**  
+Sets the default TTS (Text-to-Speech) model.
+
+**Usage**  
+```
+SET DEFAULT TTS <tts_id>;
+```
+
+**Parameters**  
+- `tts_id`: TTS model identifier, quoted string.
+
+**Example**  
+```
+ragflow> SET DEFAULT TTS 'tts-1';
+```
+
+**Display Effect**  
+(Sample output will be provided by the user)
+
+---
+
+### reset_default_llm
+
+**Description**  
+Resets the default LLM to system default.
+
+**Usage**  
+```
+RESET DEFAULT LLM;
+```
+
+**Parameters**  
+No parameters.
+
+**Example**  
+```
+ragflow> RESET DEFAULT LLM;
+```
+
+**Display Effect**  
+(Sample output will be provided by the user)
+
+---
+
+### reset_default_vlm
+
+**Description**  
+Resets the default VLM to system default.
+
+**Usage**  
+```
+RESET DEFAULT VLM;
+```
+
+**Parameters**  
+No parameters.
+
+**Example**  
+```
+ragflow> RESET DEFAULT VLM;
+```
+
+**Display Effect**  
+(Sample output will be provided by the user)
+
+---
+
+### reset_default_embedding
+
+**Description**  
+Resets the default embedding model to system default.
+
+**Usage**  
+```
+RESET DEFAULT EMBEDDING;
+```
+
+**Parameters**  
+No parameters.
+
+**Example**  
+```
+ragflow> RESET DEFAULT EMBEDDING;
+```
+
+**Display Effect**  
+(Sample output will be provided by the user)
+
+---
+
+### reset_default_reranker
+
+**Description**  
+Resets the default reranker model to system default.
+
+**Usage**  
+```
+RESET DEFAULT RERANKER;
+```
+
+**Parameters**  
+No parameters.
+
+**Example**  
+```
+ragflow> RESET DEFAULT RERANKER;
+```
+
+**Display Effect**  
+(Sample output will be provided by the user)
+
+---
+
+### reset_default_asr
+
+**Description**  
+Resets the default ASR model to system default.
+
+**Usage**  
+```
+RESET DEFAULT ASR;
+```
+
+**Parameters**  
+No parameters.
+
+**Example**  
+```
+ragflow> RESET DEFAULT ASR;
+```
+
+**Display Effect**  
+(Sample output will be provided by the user)
+
+---
+
+### reset_default_tts
+
+**Description**  
+Resets the default TTS model to system default.
+
+**Usage**  
+```
+RESET DEFAULT TTS;
+```
+
+**Parameters**  
+No parameters.
+
+**Example**  
+```
+ragflow> RESET DEFAULT TTS;
+```
+
+**Display Effect**  
+(Sample output will be provided by the user)
+
+---
+
+### create_user_dataset_with_parser
+
+**Description**  
+Creates a user dataset with the specified parser.
+
+**Usage**  
+```
+CREATE DATASET <dataset_name> WITH EMBEDDING <embedding> PARSER <parser_type>;
+```
+
+**Parameters**  
+- `dataset_name`: Dataset name, quoted string.
+- `embedding`: Embedding model name, quoted string.
+- `parser_type`: Parser type, quoted string.
+
+**Example**  
+```
+ragflow> CREATE DATASET 'my_dataset' WITH EMBEDDING 'text-embedding-ada-002' PARSER 'pdf';
+```
+
+**Display Effect**  
+(Sample output will be provided by the user)
+
+---
+
+### create_user_dataset_with_pipeline
+
+**Description**  
+Creates a user dataset with the specified pipeline.
+
+**Usage**  
+```
+CREATE DATASET <dataset_name> WITH EMBEDDING <embedding> PIPELINE <pipeline>;
+```
+
+**Parameters**  
+- `dataset_name`: Dataset name, quoted string.
+- `embedding`: Embedding model name, quoted string.
+- `pipeline`: Pipeline name, quoted string.
+
+**Example**  
+```
+ragflow> CREATE DATASET 'my_dataset' WITH EMBEDDING 'text-embedding-ada-002' PIPELINE 'standard';
+```
+
+**Display Effect**  
+(Sample output will be provided by the user)
+
+---
+
+### drop_user_dataset
+
+**Description**  
+Deletes a user dataset.
+
+**Usage**  
+```
+DROP DATASET <dataset_name>;
+```
+
+**Parameters**  
+- `dataset_name`: Name of the dataset to delete, quoted string.
+
+**Example**  
+```
+ragflow> DROP DATASET 'my_dataset';
+```
+
+**Display Effect**  
+(Sample output will be provided by the user)
+
+---
+
+### list_user_datasets
+
+**Description**  
+Lists all datasets for the current user.
+
+**Usage**  
+```
+LIST DATASETS;
+```
+
+**Parameters**  
+No parameters.
+
+**Example**  
+```
+ragflow> LIST DATASETS;
+```
+
+**Display Effect**  
+(Sample output will be provided by the user)
+
+---
+
+### list_user_dataset_files
+
+**Description**  
+Lists all files in the specified dataset.
+
+**Usage**  
+```
+LIST FILES OF DATASET <dataset_name>;
+```
+
+**Parameters**  
+- `dataset_name`: Dataset name, quoted string.
+
+**Example**  
+```
+ragflow> LIST FILES OF DATASET 'my_dataset';
+```
+
+**Display Effect**  
+(Sample output will be provided by the user)
+
+---
+
+### list_user_agents
+
+**Description**  
+Lists all agents for the current user.
+
+**Usage**  
+```
+LIST AGENTS;
+```
+
+**Parameters**  
+No parameters.
+
+**Example**  
+```
+ragflow> LIST AGENTS;
+```
+
+**Display Effect**  
+(Sample output will be provided by the user)
+
+---
+
+### list_user_chats
+
+**Description**  
+Lists all chat sessions for the current user.
+
+**Usage**  
+```
+LIST CHATS;
+```
+
+**Parameters**  
+No parameters.
+
+**Example**  
+```
+ragflow> LIST CHATS;
+```
+
+**Display Effect**  
+(Sample output will be provided by the user)
+
+---
+
+### create_user_chat
+
+**Description**  
+Creates a new chat session.
+
+**Usage**  
+```
+CREATE CHAT <chat_name>;
+```
+
+**Parameters**  
+- `chat_name`: Chat session name, quoted string.
+
+**Example**  
+```
+ragflow> CREATE CHAT 'my_chat';
+```
+
+**Display Effect**  
+(Sample output will be provided by the user)
+
+---
+
+### drop_user_chat
+
+**Description**  
+Deletes a chat session.
+
+**Usage**  
+```
+DROP CHAT <chat_name>;
+```
+
+**Parameters**  
+- `chat_name`: Name of the chat session to delete, quoted string.
+
+**Example**  
+```
+ragflow> DROP CHAT 'my_chat';
+```
+
+**Display Effect**  
+(Sample output will be provided by the user)
+
+---
+
+### list_user_model_providers
+
+**Description**  
+Lists all model providers for the current user.
+
+**Usage**  
+```
+LIST MODEL PROVIDERS;
+```
+
+**Parameters**  
+No parameters.
+
+**Example**  
+```
+ragflow> LIST MODEL PROVIDERS;
+```
+
+**Display Effect**  
+(Sample output will be provided by the user)
+
+---
+
+### list_user_default_models
+
+**Description**  
+Lists all default model settings for the current user.
+
+**Usage**  
+```
+LIST DEFAULT MODELS;
+```
+
+**Parameters**  
+No parameters.
+
+**Example**  
+```
+ragflow> LIST DEFAULT MODELS;
+```
+
+**Display Effect**  
+(Sample output will be provided by the user)
+
+---
+
+### import_docs_into_dataset
+
+**Description**  
+Imports documents into the specified dataset.
+
+**Usage**  
+```
+IMPORT <document_list> INTO DATASET <dataset_name>;
+```
+
+**Parameters**  
+- `document_list`: List of document paths, multiple paths can be separated by commas, or as a space-separated quoted string.
+- `dataset_name`: Target dataset name, quoted string.
+
+**Example**  
+```
+ragflow> IMPORT '/path/to/doc1.pdf,/path/to/doc2.pdf' INTO DATASET 'my_dataset';
+```
+
+**Display Effect**  
+(Sample output will be provided by the user)
+
+---
+
+### search_on_datasets
+
+**Description**  
+Searches in one or more specified datasets.
+
+**Usage**  
+```
+SEARCH <question> ON DATASETS <dataset_list>;
+```
+
+**Parameters**  
+- `question`: Search question, quoted string.
+- `dataset_list`: List of dataset names, multiple names can be separated by commas, or as a space-separated quoted string.
+
+**Example**  
+```
+ragflow> SEARCH 'What is RAG?' ON DATASETS 'dataset1,dataset2';
+```
+
+**Display Effect**  
+(Sample output will be provided by the user)
+
+---
+
+### parse_dataset_docs
+
+**Description**  
+Parses specified documents in a dataset.
+
+**Usage**  
+```
+PARSE <document_names> OF DATASET <dataset_name>;
+```
+
+**Parameters**  
+- `document_names`: List of document names, multiple names can be separated by commas, or as a space-separated quoted string.
+- `dataset_name`: Dataset name, quoted string.
+
+**Example**  
+```
+ragflow> PARSE 'doc1.pdf,doc2.pdf' OF DATASET 'my_dataset';
+```
+
+**Display Effect**  
+(Sample output will be provided by the user)
+
+---
+
+### parse_dataset_sync
+
+**Description**  
+Synchronously parses the entire dataset.
+
+**Usage**  
+```
+PARSE DATASET <dataset_name> SYNC;
+```
+
+**Parameters**  
+- `dataset_name`: Dataset name, quoted string.
+
+**Example**  
+```
+ragflow> PARSE DATASET 'my_dataset' SYNC;
+```
+
+**Display Effect**  
+(Sample output will be provided by the user)
+
+---
+
+### parse_dataset_async
+
+**Description**  
+Asynchronously parses the entire dataset.
+
+**Usage**  
+```
+PARSE DATASET <dataset_name> ASYNC;
+```
+
+**Parameters**  
+- `dataset_name`: Dataset name, quoted string.
+
+**Example**  
+```
+ragflow> PARSE DATASET 'my_dataset' ASYNC;
+```
+
+**Display Effect**  
+(Sample output will be provided by the user)
+
+---
+
+### benchmark
+
+**Description**  
+Performs performance benchmark testing on the specified user command.
+
+**Usage**  
+```
+BENCHMARK <concurrency> <iterations> <user_command>;
+```
+
+**Parameters**  
+- `concurrency`: Concurrency number, positive integer.
+- `iterations`: Number of iterations, positive integer.
+- `user_command`: User command to test (must be a valid user command, such as `PING;`).
+
+**Example**  
+```
+ragflow> BENCHMARK 5 10 PING;
+```
+
+**Display Effect**  
+(Sample output will be provided by the user)
+
+---
+
+**Notes**  
+- All string parameters (such as names, IDs, paths) must be enclosed in single quotes (`'`) or double quotes (`"`).
+- Commands must end with a semicolon (`;`).
+- The prompt is `ragflow>`.

admin/client/README.md

@@ -48,7 +48,7 @@ It consists of a server-side Service and a command-line client (CLI), both imple
 1.  Ensure the Admin Service is running.
 2.  Install ragflow-cli.
     ```bash
-    pip install ragflow-cli==0.24.0
+    pip install ragflow-cli==0.25.0
     ```
 3.  Launch the CLI client:
     ```bash

admin/client/parser.py

@@ -77,10 +77,17 @@ sql_command: login_user
            | drop_user_dataset
            | list_user_datasets
            | list_user_dataset_files
+           | list_user_dataset_documents
+           | list_user_datasets_metadata
+           | list_user_documents_metadata_summary
            | list_user_agents
            | list_user_chats
            | create_user_chat
            | drop_user_chat
+           | create_dataset_table
+           | drop_dataset_table
+           | create_metadata_table
+           | drop_metadata_table
            | list_user_model_providers
            | list_user_default_models
            | parse_dataset_docs
@@ -88,15 +95,35 @@ sql_command: login_user
            | parse_dataset_async
            | import_docs_into_dataset
            | search_on_datasets
+           | get_chunk
+           | list_chunks
+           | insert_dataset_from_file
+           | insert_metadata_from_file
+           | update_chunk
+           | set_metadata
+           | remove_tags
+           | remove_chunks
+           | create_chat_session
+           | drop_chat_session
+           | list_chat_sessions
+           | chat_on_session
+           | list_server_configs
+           | show_fingerprint
+           | set_license
+           | set_license_config
+           | show_license
+           | check_license
            | benchmark
 
 // meta command definition
 meta_command: "\\" meta_command_name [meta_args]
 
+COMMA: ","
+
 meta_command_name: /[a-zA-Z?]+/
 meta_args: (meta_arg)+
 
-meta_arg: /[^\\s"']+/ | quoted_string
+meta_arg: /[^\s"',]+/ | quoted_string
 
 // command definition
 
@@ -117,6 +144,7 @@ ALTER: "ALTER"i
 ACTIVE: "ACTIVE"i
 ADMIN: "ADMIN"i
 PASSWORD: "PASSWORD"i
+DATASET_TABLE: "DATASET TABLE"i
 DATASET: "DATASET"i
 DATASETS: "DATASETS"i
 OF: "OF"i
@@ -151,11 +179,18 @@ DEFAULT: "DEFAULT"i
 CHATS: "CHATS"i
 CHAT: "CHAT"i
 FILES: "FILES"i
+DOCUMENT: "DOCUMENT"i
+DOCUMENTS: "DOCUMENTS"i
+METADATA: "METADATA"i
+SUMMARY: "SUMMARY"i
 AS: "AS"i
 PARSE: "PARSE"i
 IMPORT: "IMPORT"i
 INTO: "INTO"i
+IN: "IN"i
 WITH: "WITH"i
+VECTOR: "VECTOR"i
+SIZE: "SIZE"i
 PARSER: "PARSER"i
 PIPELINE: "PIPELINE"i
 SEARCH: "SEARCH"i
@@ -170,8 +205,28 @@ ASYNC: "ASYNC"i
 SYNC: "SYNC"i
 BENCHMARK: "BENCHMARK"i
 PING: "PING"i
+SESSION: "SESSION"i
+SESSIONS: "SESSIONS"i
+SERVER: "SERVER"i
+FINGERPRINT: "FINGERPRINT"i
+LICENSE: "LICENSE"i
+CHECK: "CHECK"i
+CONFIG: "CONFIG"i
+INDEX: "INDEX"i
+TABLE: "TABLE"i
+CHUNK: "CHUNK"i
+CHUNKS: "CHUNKS"i
+GET: "GET"i
+INSERT: "INSERT"i
+PAGE: "PAGE"i
+KEYWORDS: "KEYWORDS"i
+AVAILABLE: "AVAILABLE"i
+FILE: "FILE"i
+UPDATE: "UPDATE"i
+REMOVE: "REMOVE"i
+TAGS: "TAGS"i
 
-login_user: LOGIN USER quoted_string ";"
+login_user: LOGIN USER quoted_string (PASSWORD quoted_string)? ";"
 list_services: LIST SERVICES ";"
 show_service: SHOW SERVICE NUMBER ";"
 startup_service: STARTUP SERVICE NUMBER ";"
@@ -215,6 +270,14 @@ list_variables: LIST VARS ";"
 list_configs: LIST CONFIGS ";"
 list_environments: LIST ENVS ";"
 
+show_fingerprint: SHOW FINGERPRINT ";"
+set_license: SET LICENSE quoted_string ";"
+set_license_config: SET LICENSE CONFIG NUMBER NUMBER ";"
+show_license: SHOW LICENSE ";"
+check_license: CHECK LICENSE ";"
+
+list_server_configs: LIST SERVER CONFIGS ";"
+
 benchmark: BENCHMARK NUMBER NUMBER user_statement
 
 user_statement: ping_server
@@ -246,6 +309,13 @@ user_statement: ping_server
                 | list_user_default_models
                 | import_docs_into_dataset
                 | search_on_datasets
+                | update_chunk
+                | set_metadata
+                | remove_tags
+                | create_chat_session
+                | drop_chat_session
+                | list_chat_sessions
+                | chat_on_session
 
 ping_server: PING ";"
 show_current_user: SHOW CURRENT USER ";"
@@ -270,24 +340,46 @@ create_user_dataset_with_parser: CREATE DATASET quoted_string WITH EMBEDDING quo
 create_user_dataset_with_pipeline: CREATE DATASET quoted_string WITH EMBEDDING quoted_string PIPELINE quoted_string ";" 
 drop_user_dataset: DROP DATASET quoted_string ";"
 list_user_dataset_files: LIST FILES OF DATASET quoted_string ";"
+list_user_dataset_documents: LIST DOCUMENTS OF DATASET quoted_string ";"
+list_user_datasets_metadata: LIST METADATA OF DATASETS quoted_string (COMMA quoted_string)* ";"
+list_user_documents_metadata_summary: LIST METADATA SUMMARY OF DATASET quoted_string (DOCUMENTS quoted_string (COMMA quoted_string)*)? ";"
 list_user_agents: LIST AGENTS ";"
 list_user_chats: LIST CHATS ";"
 create_user_chat: CREATE CHAT quoted_string ";"
 drop_user_chat: DROP CHAT quoted_string ";"
+create_chat_session: CREATE CHAT quoted_string SESSION ";"
+drop_chat_session: DROP CHAT quoted_string SESSION quoted_string ";"
+list_chat_sessions: LIST CHAT quoted_string SESSIONS ";"
+chat_on_session: CHAT quoted_string ON quoted_string SESSION quoted_string ";"
 list_user_model_providers: LIST MODEL PROVIDERS ";"
 list_user_default_models: LIST DEFAULT MODELS ";"
 import_docs_into_dataset: IMPORT quoted_string INTO DATASET quoted_string ";"
 search_on_datasets: SEARCH quoted_string ON DATASETS quoted_string ";"
+get_chunk: GET CHUNK quoted_string ";"
+list_chunks: LIST CHUNKS OF DOCUMENT quoted_string ("PAGE" NUMBER)? ("SIZE" NUMBER)? ("KEYWORDS" quoted_string)? ("AVAILABLE" NUMBER)? ";"
+set_metadata: SET METADATA OF DOCUMENT quoted_string TO quoted_string ";"
+remove_tags: REMOVE TAGS quoted_string (COMMA quoted_string)* FROM DATASET quoted_string ";"
+remove_chunks: REMOVE CHUNKS quoted_string (COMMA quoted_string)* FROM DOCUMENT quoted_string ";"
+           | REMOVE ALL CHUNKS FROM DOCUMENT quoted_string ";"
 
 parse_dataset_docs: PARSE quoted_string OF DATASET quoted_string ";"
 parse_dataset_sync: PARSE DATASET quoted_string SYNC ";"
 parse_dataset_async: PARSE DATASET quoted_string ASYNC ";"
 
-identifier_list: identifier ("," identifier)*
+// Internal CLI only for GO
+create_dataset_table: CREATE DATASET TABLE quoted_string VECTOR SIZE NUMBER ";"
+drop_dataset_table: DROP DATASET TABLE quoted_string ";"
+create_metadata_table: CREATE METADATA TABLE ";"
+drop_metadata_table: DROP METADATA TABLE ";"
+insert_dataset_from_file: INSERT DATASET FROM FILE quoted_string ";"
+insert_metadata_from_file: INSERT METADATA FROM FILE quoted_string ";"
+update_chunk: UPDATE CHUNK quoted_string OF DATASET quoted_string SET quoted_string ";"
+
+identifier_list: identifier (COMMA identifier)*
 
 identifier: WORD
 quoted_string: QUOTED_STRING
-status: WORD
+status: ON | WORD
 
 QUOTED_STRING: /'[^']+'/ | /"[^"]+"/
 WORD: /[a-zA-Z0-9_\-\.]+/
@@ -307,7 +399,13 @@ class RAGFlowCLITransformer(Transformer):
 
     def login_user(self, items):
         email = items[2].children[0].strip("'\"")
-        return {"type": "login_user", "email": email}
+        if len(items) == 5:
+            # With password: LOGIN USER email PASSWORD password
+            password = items[4].children[0].strip("'\"")
+            return {"type": "login_user", "email": email, "password": password}
+        else:
+            # Without password: LOGIN USER email
+            return {"type": "login_user", "email": email}
 
     def ping_server(self, items):
         return {"type": "ping_server"}
@@ -459,6 +557,27 @@ class RAGFlowCLITransformer(Transformer):
     def list_environments(self, items):
         return {"type": "list_environments"}
 
+    def show_fingerprint(self, items):
+        return {"type": "show_fingerprint"}
+
+    def set_license(self, items):
+        license = items[2].children[0].strip("'\"")
+        return {"type": "set_license", "license": license}
+
+    def set_license_config(self, items):
+        value1: int = int(items[3])
+        value2: int = int(items[4])
+        return {"type": "set_license_config", "value1": value1, "value2": value2}
+
+    def show_license(self, items):
+        return {"type": "show_license"}
+
+    def check_license(self, items):
+        return {"type": "check_license"}
+
+    def list_server_configs(self, items):
+        return {"type": "list_server_configs"}
+
     def create_model_provider(self, items):
         provider_name = items[3].children[0].strip("'\"")
         provider_key = items[4].children[0].strip("'\"")
@@ -538,6 +657,28 @@ class RAGFlowCLITransformer(Transformer):
         dataset_name = items[4].children[0].strip("'\"")
         return {"type": "list_user_dataset_files", "dataset_name": dataset_name}
 
+    def list_user_dataset_documents(self, items):
+        dataset_name = items[4].children[0].strip("'\"")
+        return {"type": "list_user_dataset_documents", "dataset_name": dataset_name}
+
+    def list_user_datasets_metadata(self, items):
+        dataset_names = []
+        dataset_names.append(items[4].children[0].strip("'\""))
+        for i in range(5, len(items)):
+            if items[i] and hasattr(items[i], 'children') and items[i].children:
+                dataset_names.append(items[i].children[0].strip("'\""))
+        return {"type": "list_user_datasets_metadata", "dataset_names": dataset_names}
+
+    def list_user_documents_metadata_summary(self, items):
+        dataset_name = items[5].children[0].strip("'\"")
+        doc_ids = []
+        if len(items) > 6 and items[6] == "DOCUMENTS":
+            for i in range(7, len(items)):
+                if items[i] and hasattr(items[i], 'children') and items[i].children:
+                    doc_id = items[i].children[0].strip("'\"")
+                    doc_ids.append(doc_id)
+        return {"type": "list_user_documents_metadata_summary", "dataset_name": dataset_name, "document_ids": doc_ids}
+
     def list_user_agents(self, items):
         return {"type": "list_user_agents"}
 
@@ -552,6 +693,30 @@ class RAGFlowCLITransformer(Transformer):
         chat_name = items[2].children[0].strip("'\"")
         return {"type": "drop_user_chat", "chat_name": chat_name}
 
+    def create_dataset_table(self, items):
+        dataset_name = None
+        vector_size = None
+        for i, item in enumerate(items):
+            if hasattr(item, 'data') and item.data == 'quoted_string':
+                dataset_name = item.children[0].strip("'\"")
+            if hasattr(item, 'type') and item.type == 'NUMBER':
+                if i > 0 and items[i-1].type == 'SIZE' and items[i-2].type == 'VECTOR':
+                    vector_size = int(item)
+        return {"type": "create_dataset_table", "dataset_name": dataset_name, "vector_size": vector_size}
+
+    def drop_dataset_table(self, items):
+        dataset_name = None
+        for item in items:
+            if hasattr(item, 'data') and item.data == 'quoted_string':
+                dataset_name = item.children[0].strip("'\"")
+        return {"type": "drop_dataset_table", "dataset_name": dataset_name}
+
+    def create_metadata_table(self, items):
+        return {"type": "create_metadata_table"}
+
+    def drop_metadata_table(self, items):
+        return {"type": "drop_metadata_table"}
+
     def list_user_model_providers(self, items):
         return {"type": "list_user_model_providers"}
 
@@ -575,6 +740,25 @@ class RAGFlowCLITransformer(Transformer):
         dataset_name = items[2].children[0].strip("'\"")
         return {"type": "parse_dataset", "dataset_name": dataset_name, "method": "async"}
 
+    def create_chat_session(self, items):
+        chat_name = items[2].children[0].strip("'\"")
+        return {"type": "create_chat_session", "chat_name": chat_name}
+
+    def drop_chat_session(self, items):
+        chat_name = items[2].children[0].strip("'\"")
+        session_id = items[4].children[0].strip("'\"")
+        return {"type": "drop_chat_session", "chat_name": chat_name, "session_id": session_id}
+
+    def list_chat_sessions(self, items):
+        chat_name = items[2].children[0].strip("'\"")
+        return {"type": "list_chat_sessions", "chat_name": chat_name}
+
+    def chat_on_session(self, items):
+        message = items[1].children[0].strip("'\"")
+        chat_name = items[3].children[0].strip("'\"")
+        session_id = items[5].children[0].strip("'\"")
+        return {"type": "chat_on_session", "message": message, "chat_name": chat_name, "session_id": session_id}
+
     def import_docs_into_dataset(self, items):
         document_list_str = items[1].children[0].strip("'\"")
         document_paths = document_list_str.split(",")
@@ -593,6 +777,103 @@ class RAGFlowCLITransformer(Transformer):
             datasets = datasets.split(" ")
         return {"type": "search_on_datasets", "datasets": datasets, "question": question}
 
+    def get_chunk(self, items):
+        chunk_id = items[2].children[0].strip("'\"")
+        return {"type": "get_chunk", "chunk_id": chunk_id}
+
+    def insert_dataset_from_file(self, items):
+        file_path = items[4].children[0].strip("'\"")
+        return {"type": "insert_dataset_from_file", "file_path": file_path}
+
+    def insert_metadata_from_file(self, items):
+        file_path = items[4].children[0].strip("'\"")
+        return {"type": "insert_metadata_from_file", "file_path": file_path}
+
+    def update_chunk(self, items):
+        def get_quoted_value(item):
+            if hasattr(item, 'children') and item.children:
+                return item.children[0].strip("'\"")
+            return str(item).strip("'\"")
+
+        chunk_id = get_quoted_value(items[2])
+        dataset_name = get_quoted_value(items[5])
+        json_body = get_quoted_value(items[7])
+        return {"type": "update_chunk", "chunk_id": chunk_id, "dataset_name": dataset_name, "json_body": json_body}
+
+    def set_metadata(self, items):
+        doc_id = items[4].children[0].strip("'\"")
+        meta_json = items[6].children[0].strip("'\"")
+        return {"type": "set_metadata", "doc_id": doc_id, "meta": meta_json}
+
+    def remove_tags(self, items):
+        # items: REMOVE, TAGS, quoted_string(tag1), quoted_string(tag2), ..., FROM, DATASET, quoted_string(dataset_name), ";"
+        tags = []
+        # Start from index 2 (after TAGS keyword) and parse quoted strings until FROM
+        for i in range(2, len(items)):
+            item = items[i]
+            # Check for FROM token to stop
+            if hasattr(item, 'type') and item.type == 'FROM':
+                break
+            if hasattr(item, 'children') and item.children:
+                tag = item.children[0].strip("'\"")
+                tags.append(tag)
+        # Find dataset_name: quoted_string after DATASET
+        dataset_name = None
+        for i, item in enumerate(items):
+            # Check if item is a DATASET token
+            if hasattr(item, 'type') and item.type == 'DATASET':
+                # Next item should be quoted_string
+                dataset_name = items[i + 1].children[0].strip("'\"")
+                break
+        return {"type": "remove_tags", "dataset_name": dataset_name, "tags": tags}
+
+    def remove_chunks(self, items):
+        # Handle two cases:
+        # 1. REMOVE CHUNKS quoted_string (COMMA quoted_string)* FROM DOCUMENT quoted_string ";"
+        # 2. REMOVE ALL CHUNKS FROM DOCUMENT quoted_string ";"
+
+        # Check if it's "REMOVE ALL CHUNKS"
+        for item in items:
+            if hasattr(item, 'type') and item.type == 'ALL':
+                # Find doc_id
+                for j, inner_item in enumerate(items):
+                    if hasattr(inner_item, 'type') and inner_item.type == 'DOCUMENT':
+                        doc_id = items[j + 1].children[0].strip("'\"")
+                        return {"type": "remove_chunks", "doc_id": doc_id, "delete_all": True}
+
+        # Otherwise, we have chunk_ids
+        chunk_ids = []
+        doc_id = None
+        for i, item in enumerate(items):
+            if hasattr(item, 'type') and item.type == 'DOCUMENT':
+                doc_id = items[i + 1].children[0].strip("'\"")
+            elif hasattr(item, 'children') and item.children:
+                val = item.children[0].strip("'\"")
+                # Skip if it's "FROM" or "DOCUMENT"
+                if val.upper() in ['FROM', 'DOCUMENT']:
+                    continue
+                chunk_ids.append(val)
+
+        return {"type": "remove_chunks", "doc_id": doc_id, "chunk_ids": chunk_ids}
+
+    def list_chunks(self, items):
+        doc_id = items[4].children[0].strip("'\"")
+        result = {"type": "list_chunks", "doc_id": doc_id}
+
+        # Parse optional parameters: PAGE, SIZE, KEYWORDS, AVAILABLE
+        # items structure varies based on which params are present
+        for i, item in enumerate(items):
+            if str(item) == "PAGE":
+                result["page"] = int(items[i + 1])
+            elif str(item) == "SIZE":
+                result["size"] = int(items[i + 1])
+            elif str(item) == "KEYWORDS":
+                result["keywords"] = items[i + 1].children[0].strip("'\"")
+            elif str(item) == "AVAILABLE":
+                result["available_int"] = int(items[i + 1])
+
+        return result
+
     def benchmark(self, items):
         concurrency: int = int(items[1])
         iterations: int = int(items[2])

admin/client/pyproject.toml

@@ -1,6 +1,6 @@
 [project]
 name = "ragflow-cli"
-version = "0.24.0"
+version = "0.25.0"
 description = "Admin Service's client of [RAGFlow](https://github.com/infiniflow/ragflow). The Admin Service provides user management and system monitoring. "
 authors = [{ name = "Lynn", email = "lynn_inf@hotmail.com" }]
 license = { text = "Apache License, Version 2.0" }
@@ -11,17 +11,17 @@ dependencies = [
     "beartype>=0.20.0,<1.0.0",
     "pycryptodomex>=3.10.0",
     "lark>=1.1.0",
+    "requests-toolbelt>=1.0.0",
 ]
 
 [dependency-groups]
 test = [
     "pytest>=8.3.5",
     "requests>=2.32.3",
-    "requests-toolbelt>=1.0.0",
 ]
 
 [tool.setuptools]
-py-modules = ["ragflow_cli", "parser"]
+py-modules = ["ragflow_cli", "parser", "http_client", "ragflow_client", "user"]
 
 [project.scripts]
 ragflow-cli = "ragflow_cli:main"

admin/client/ragflow_cli.py

@@ -18,6 +18,9 @@ import sys
 import argparse
 import base64
 import getpass
+import os
+import atexit
+import readline
 from cmd import Cmd
 from typing import Any, Dict, List
 
@@ -61,6 +64,12 @@ class RAGFlowCLI(Cmd):
         self.port: int = 0
         self.mode: str = "admin"
         self.ragflow_client = None
+        # History file for readline persistence
+        self.history_file = os.path.expanduser("~/.ragflow_cli_history")
+        # Load existing history
+        self._load_history()
+        # Register cleanup to save history on exit
+        atexit.register(self._save_history)
 
     intro = r"""Type "\h" for help."""
     prompt = "ragflow> "
@@ -99,6 +108,7 @@ class RAGFlowCLI(Cmd):
             return {"type": "empty"}
 
         self.command_history.append(command_str)
+        readline.add_history(command_str)
 
         try:
             result = self.parser.parse(command_str)
@@ -210,6 +220,21 @@ class RAGFlowCLI(Cmd):
 
         print(separator)
 
+    def _load_history(self):
+        """Load command history from file."""
+        try:
+            if os.path.exists(self.history_file):
+                readline.read_history_file(self.history_file)
+        except Exception:
+            pass  # Ignore errors loading history
+
+    def _save_history(self):
+        """Save command history to file."""
+        try:
+            readline.write_history_file(self.history_file)
+        except Exception:
+            pass  # Ignore errors saving history
+
     def run_interactive(self, args):
         if self.verify_auth(args, single_command=False, auth=args["auth"]):
             print(r"""

admin/client/user.py

@@ -26,7 +26,19 @@ class AuthException(Exception):
 
 def encrypt_password(password_plain: str) -> str:
     try:
-        from api.utils.crypt import crypt
+        import base64
+        from Cryptodome.PublicKey import RSA
+        from Cryptodome.Cipher import PKCS1_v1_5 as Cipher_pkcs1_v1_5
+        def crypt(line):
+            """
+            decrypt(crypt(input_string)) == base64(input_string), which frontend and ragflow_cli use.
+            """
+            pub = "-----BEGIN PUBLIC KEY-----\nMIIBIjANBgkqhkiG9w0BAQEFAAOCAQ8AMIIBCgKCAQEArq9XTUSeYr2+N1h3Afl/z8Dse/2yD0ZGrKwx+EEEcdsBLca9Ynmx3nIB5obmLlSfmskLpBo0UACBmB5rEjBp2Q2f3AG3Hjd4B+gNCG6BDaawuDlgANIhGnaTLrIqWrrcm4EMzJOnAOI1fgzJRsOOUEfaS318Eq9OVO3apEyCCt0lOQK6PuksduOjVxtltDav+guVAA068NrPYmRNabVKRNLJpL8w4D44sfth5RvZ3q9t+6RTArpEtc5sh5ChzvqPOzKGMXW83C95TxmXqpbK6olN4RevSfVjEAgCydH6HN6OhtOQEcnrU97r9H0iZOWwbw3pVrZiUkuRD1R56Wzs2wIDAQAB\n-----END PUBLIC KEY-----"
+            rsa_key = RSA.importKey(pub)
+            cipher = Cipher_pkcs1_v1_5.new(rsa_key)
+            password_base64 = base64.b64encode(line.encode('utf-8')).decode("utf-8")
+            encrypted_password = cipher.encrypt(password_base64.encode())
+            return base64.b64encode(encrypted_password).decode('utf-8')
     except Exception as exc:
         raise AuthException(
             "Password encryption unavailable; install pycryptodomex (uv sync --python 3.12 --group test)."

admin/client/uv.lock

@@ -1,6 +1,6 @@
 version = 1
 revision = 3
-requires-python = ">=3.10, <3.13"
+requires-python = ">=3.12, <3.15"
 
 [[package]]
 name = "beartype"
@@ -26,38 +26,6 @@ version = "3.4.4"
 source = { registry = "https://pypi.tuna.tsinghua.edu.cn/simple" }
 sdist = { url = "https://pypi.tuna.tsinghua.edu.cn/packages/13/69/33ddede1939fdd074bce5434295f38fae7136463422fe4fd3e0e89b98062/charset_normalizer-3.4.4.tar.gz", hash = "sha256:94537985111c35f28720e43603b8e7b43a6ecfb2ce1d3058bbe955b73404e21a", size = 129418, upload-time = "2025-10-14T04:42:32.879Z" }
 wheels = [
-    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/1f/b8/6d51fc1d52cbd52cd4ccedd5b5b2f0f6a11bbf6765c782298b0f3e808541/charset_normalizer-3.4.4-cp310-cp310-macosx_10_9_universal2.whl", hash = "sha256:e824f1492727fa856dd6eda4f7cee25f8518a12f3c4a56a74e8095695089cf6d", size = 209709, upload-time = "2025-10-14T04:40:11.385Z" },
-    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/5c/af/1f9d7f7faafe2ddfb6f72a2e07a548a629c61ad510fe60f9630309908fef/charset_normalizer-3.4.4-cp310-cp310-manylinux2014_aarch64.manylinux_2_17_aarch64.manylinux_2_28_aarch64.whl", hash = "sha256:4bd5d4137d500351a30687c2d3971758aac9a19208fc110ccb9d7188fbe709e8", size = 148814, upload-time = "2025-10-14T04:40:13.135Z" },
-    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/79/3d/f2e3ac2bbc056ca0c204298ea4e3d9db9b4afe437812638759db2c976b5f/charset_normalizer-3.4.4-cp310-cp310-manylinux2014_armv7l.manylinux_2_17_armv7l.manylinux_2_31_armv7l.whl", hash = "sha256:027f6de494925c0ab2a55eab46ae5129951638a49a34d87f4c3eda90f696b4ad", size = 144467, upload-time = "2025-10-14T04:40:14.728Z" },
-    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/ec/85/1bf997003815e60d57de7bd972c57dc6950446a3e4ccac43bc3070721856/charset_normalizer-3.4.4-cp310-cp310-manylinux2014_ppc64le.manylinux_2_17_ppc64le.manylinux_2_28_ppc64le.whl", hash = "sha256:f820802628d2694cb7e56db99213f930856014862f3fd943d290ea8438d07ca8", size = 162280, upload-time = "2025-10-14T04:40:16.14Z" },
-    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/3e/8e/6aa1952f56b192f54921c436b87f2aaf7c7a7c3d0d1a765547d64fd83c13/charset_normalizer-3.4.4-cp310-cp310-manylinux2014_s390x.manylinux_2_17_s390x.manylinux_2_28_s390x.whl", hash = "sha256:798d75d81754988d2565bff1b97ba5a44411867c0cf32b77a7e8f8d84796b10d", size = 159454, upload-time = "2025-10-14T04:40:17.567Z" },
-    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/36/3b/60cbd1f8e93aa25d1c669c649b7a655b0b5fb4c571858910ea9332678558/charset_normalizer-3.4.4-cp310-cp310-manylinux2014_x86_64.manylinux_2_17_x86_64.manylinux_2_28_x86_64.whl", hash = "sha256:9d1bb833febdff5c8927f922386db610b49db6e0d4f4ee29601d71e7c2694313", size = 153609, upload-time = "2025-10-14T04:40:19.08Z" },
-    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/64/91/6a13396948b8fd3c4b4fd5bc74d045f5637d78c9675585e8e9fbe5636554/charset_normalizer-3.4.4-cp310-cp310-manylinux_2_31_riscv64.manylinux_2_39_riscv64.whl", hash = "sha256:9cd98cdc06614a2f768d2b7286d66805f94c48cde050acdbbb7db2600ab3197e", size = 151849, upload-time = "2025-10-14T04:40:20.607Z" },
-    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/b7/7a/59482e28b9981d105691e968c544cc0df3b7d6133152fb3dcdc8f135da7a/charset_normalizer-3.4.4-cp310-cp310-musllinux_1_2_aarch64.whl", hash = "sha256:077fbb858e903c73f6c9db43374fd213b0b6a778106bc7032446a8e8b5b38b93", size = 151586, upload-time = "2025-10-14T04:40:21.719Z" },
-    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/92/59/f64ef6a1c4bdd2baf892b04cd78792ed8684fbc48d4c2afe467d96b4df57/charset_normalizer-3.4.4-cp310-cp310-musllinux_1_2_armv7l.whl", hash = "sha256:244bfb999c71b35de57821b8ea746b24e863398194a4014e4c76adc2bbdfeff0", size = 145290, upload-time = "2025-10-14T04:40:23.069Z" },
-    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/6b/63/3bf9f279ddfa641ffa1962b0db6a57a9c294361cc2f5fcac997049a00e9c/charset_normalizer-3.4.4-cp310-cp310-musllinux_1_2_ppc64le.whl", hash = "sha256:64b55f9dce520635f018f907ff1b0df1fdc31f2795a922fb49dd14fbcdf48c84", size = 163663, upload-time = "2025-10-14T04:40:24.17Z" },
-    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/ed/09/c9e38fc8fa9e0849b172b581fd9803bdf6e694041127933934184e19f8c3/charset_normalizer-3.4.4-cp310-cp310-musllinux_1_2_riscv64.whl", hash = "sha256:faa3a41b2b66b6e50f84ae4a68c64fcd0c44355741c6374813a800cd6695db9e", size = 151964, upload-time = "2025-10-14T04:40:25.368Z" },
-    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/d2/d1/d28b747e512d0da79d8b6a1ac18b7ab2ecfd81b2944c4c710e166d8dd09c/charset_normalizer-3.4.4-cp310-cp310-musllinux_1_2_s390x.whl", hash = "sha256:6515f3182dbe4ea06ced2d9e8666d97b46ef4c75e326b79bb624110f122551db", size = 161064, upload-time = "2025-10-14T04:40:26.806Z" },
-    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/bb/9a/31d62b611d901c3b9e5500c36aab0ff5eb442043fb3a1c254200d3d397d9/charset_normalizer-3.4.4-cp310-cp310-musllinux_1_2_x86_64.whl", hash = "sha256:cc00f04ed596e9dc0da42ed17ac5e596c6ccba999ba6bd92b0e0aef2f170f2d6", size = 155015, upload-time = "2025-10-14T04:40:28.284Z" },
-    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/1f/f3/107e008fa2bff0c8b9319584174418e5e5285fef32f79d8ee6a430d0039c/charset_normalizer-3.4.4-cp310-cp310-win32.whl", hash = "sha256:f34be2938726fc13801220747472850852fe6b1ea75869a048d6f896838c896f", size = 99792, upload-time = "2025-10-14T04:40:29.613Z" },
-    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/eb/66/e396e8a408843337d7315bab30dbf106c38966f1819f123257f5520f8a96/charset_normalizer-3.4.4-cp310-cp310-win_amd64.whl", hash = "sha256:a61900df84c667873b292c3de315a786dd8dac506704dea57bc957bd31e22c7d", size = 107198, upload-time = "2025-10-14T04:40:30.644Z" },
-    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/b5/58/01b4f815bf0312704c267f2ccb6e5d42bcc7752340cd487bc9f8c3710597/charset_normalizer-3.4.4-cp310-cp310-win_arm64.whl", hash = "sha256:cead0978fc57397645f12578bfd2d5ea9138ea0fac82b2f63f7f7c6877986a69", size = 100262, upload-time = "2025-10-14T04:40:32.108Z" },
-    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/ed/27/c6491ff4954e58a10f69ad90aca8a1b6fe9c5d3c6f380907af3c37435b59/charset_normalizer-3.4.4-cp311-cp311-macosx_10_9_universal2.whl", hash = "sha256:6e1fcf0720908f200cd21aa4e6750a48ff6ce4afe7ff5a79a90d5ed8a08296f8", size = 206988, upload-time = "2025-10-14T04:40:33.79Z" },
-    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/94/59/2e87300fe67ab820b5428580a53cad894272dbb97f38a7a814a2a1ac1011/charset_normalizer-3.4.4-cp311-cp311-manylinux2014_aarch64.manylinux_2_17_aarch64.manylinux_2_28_aarch64.whl", hash = "sha256:5f819d5fe9234f9f82d75bdfa9aef3a3d72c4d24a6e57aeaebba32a704553aa0", size = 147324, upload-time = "2025-10-14T04:40:34.961Z" },
-    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/07/fb/0cf61dc84b2b088391830f6274cb57c82e4da8bbc2efeac8c025edb88772/charset_normalizer-3.4.4-cp311-cp311-manylinux2014_armv7l.manylinux_2_17_armv7l.manylinux_2_31_armv7l.whl", hash = "sha256:a59cb51917aa591b1c4e6a43c132f0cdc3c76dbad6155df4e28ee626cc77a0a3", size = 142742, upload-time = "2025-10-14T04:40:36.105Z" },
-    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/62/8b/171935adf2312cd745d290ed93cf16cf0dfe320863ab7cbeeae1dcd6535f/charset_normalizer-3.4.4-cp311-cp311-manylinux2014_ppc64le.manylinux_2_17_ppc64le.manylinux_2_28_ppc64le.whl", hash = "sha256:8ef3c867360f88ac904fd3f5e1f902f13307af9052646963ee08ff4f131adafc", size = 160863, upload-time = "2025-10-14T04:40:37.188Z" },
-    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/09/73/ad875b192bda14f2173bfc1bc9a55e009808484a4b256748d931b6948442/charset_normalizer-3.4.4-cp311-cp311-manylinux2014_s390x.manylinux_2_17_s390x.manylinux_2_28_s390x.whl", hash = "sha256:d9e45d7faa48ee908174d8fe84854479ef838fc6a705c9315372eacbc2f02897", size = 157837, upload-time = "2025-10-14T04:40:38.435Z" },
-    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/6d/fc/de9cce525b2c5b94b47c70a4b4fb19f871b24995c728e957ee68ab1671ea/charset_normalizer-3.4.4-cp311-cp311-manylinux2014_x86_64.manylinux_2_17_x86_64.manylinux_2_28_x86_64.whl", hash = "sha256:840c25fb618a231545cbab0564a799f101b63b9901f2569faecd6b222ac72381", size = 151550, upload-time = "2025-10-14T04:40:40.053Z" },
-    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/55/c2/43edd615fdfba8c6f2dfbd459b25a6b3b551f24ea21981e23fb768503ce1/charset_normalizer-3.4.4-cp311-cp311-manylinux_2_31_riscv64.manylinux_2_39_riscv64.whl", hash = "sha256:ca5862d5b3928c4940729dacc329aa9102900382fea192fc5e52eb69d6093815", size = 149162, upload-time = "2025-10-14T04:40:41.163Z" },
-    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/03/86/bde4ad8b4d0e9429a4e82c1e8f5c659993a9a863ad62c7df05cf7b678d75/charset_normalizer-3.4.4-cp311-cp311-musllinux_1_2_aarch64.whl", hash = "sha256:d9c7f57c3d666a53421049053eaacdd14bbd0a528e2186fcb2e672effd053bb0", size = 150019, upload-time = "2025-10-14T04:40:42.276Z" },
-    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/1f/86/a151eb2af293a7e7bac3a739b81072585ce36ccfb4493039f49f1d3cae8c/charset_normalizer-3.4.4-cp311-cp311-musllinux_1_2_armv7l.whl", hash = "sha256:277e970e750505ed74c832b4bf75dac7476262ee2a013f5574dd49075879e161", size = 143310, upload-time = "2025-10-14T04:40:43.439Z" },
-    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/b5/fe/43dae6144a7e07b87478fdfc4dbe9efd5defb0e7ec29f5f58a55aeef7bf7/charset_normalizer-3.4.4-cp311-cp311-musllinux_1_2_ppc64le.whl", hash = "sha256:31fd66405eaf47bb62e8cd575dc621c56c668f27d46a61d975a249930dd5e2a4", size = 162022, upload-time = "2025-10-14T04:40:44.547Z" },
-    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/80/e6/7aab83774f5d2bca81f42ac58d04caf44f0cc2b65fc6db2b3b2e8a05f3b3/charset_normalizer-3.4.4-cp311-cp311-musllinux_1_2_riscv64.whl", hash = "sha256:0d3d8f15c07f86e9ff82319b3d9ef6f4bf907608f53fe9d92b28ea9ae3d1fd89", size = 149383, upload-time = "2025-10-14T04:40:46.018Z" },
-    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/4f/e8/b289173b4edae05c0dde07f69f8db476a0b511eac556dfe0d6bda3c43384/charset_normalizer-3.4.4-cp311-cp311-musllinux_1_2_s390x.whl", hash = "sha256:9f7fcd74d410a36883701fafa2482a6af2ff5ba96b9a620e9e0721e28ead5569", size = 159098, upload-time = "2025-10-14T04:40:47.081Z" },
-    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/d8/df/fe699727754cae3f8478493c7f45f777b17c3ef0600e28abfec8619eb49c/charset_normalizer-3.4.4-cp311-cp311-musllinux_1_2_x86_64.whl", hash = "sha256:ebf3e58c7ec8a8bed6d66a75d7fb37b55e5015b03ceae72a8e7c74495551e224", size = 152991, upload-time = "2025-10-14T04:40:48.246Z" },
-    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/1a/86/584869fe4ddb6ffa3bd9f491b87a01568797fb9bd8933f557dba9771beaf/charset_normalizer-3.4.4-cp311-cp311-win32.whl", hash = "sha256:eecbc200c7fd5ddb9a7f16c7decb07b566c29fa2161a16cf67b8d068bd21690a", size = 99456, upload-time = "2025-10-14T04:40:49.376Z" },
-    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/65/f6/62fdd5feb60530f50f7e38b4f6a1d5203f4d16ff4f9f0952962c044e919a/charset_normalizer-3.4.4-cp311-cp311-win_amd64.whl", hash = "sha256:5ae497466c7901d54b639cf42d5b8c1b6a4fead55215500d2f486d34db48d016", size = 106978, upload-time = "2025-10-14T04:40:50.844Z" },
-    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/7a/9d/0710916e6c82948b3be62d9d398cb4fcf4e97b56d6a6aeccd66c4b2f2bd5/charset_normalizer-3.4.4-cp311-cp311-win_arm64.whl", hash = "sha256:65e2befcd84bc6f37095f5961e68a6f077bf44946771354a28ad434c2cce0ae1", size = 99969, upload-time = "2025-10-14T04:40:52.272Z" },
     { url = "https://pypi.tuna.tsinghua.edu.cn/packages/f3/85/1637cd4af66fa687396e757dec650f28025f2a2f5a5531a3208dc0ec43f2/charset_normalizer-3.4.4-cp312-cp312-macosx_10_13_universal2.whl", hash = "sha256:0a98e6759f854bd25a58a73fa88833fba3b7c491169f86ce1180c948ab3fd394", size = 208425, upload-time = "2025-10-14T04:40:53.353Z" },
     { url = "https://pypi.tuna.tsinghua.edu.cn/packages/9d/6a/04130023fef2a0d9c62d0bae2649b69f7b7d8d24ea5536feef50551029df/charset_normalizer-3.4.4-cp312-cp312-manylinux2014_aarch64.manylinux_2_17_aarch64.manylinux_2_28_aarch64.whl", hash = "sha256:b5b290ccc2a263e8d185130284f8501e3e36c5e02750fc6b6bdeb2e9e96f1e25", size = 148162, upload-time = "2025-10-14T04:40:54.558Z" },
     { url = "https://pypi.tuna.tsinghua.edu.cn/packages/78/29/62328d79aa60da22c9e0b9a66539feae06ca0f5a4171ac4f7dc285b83688/charset_normalizer-3.4.4-cp312-cp312-manylinux2014_armv7l.manylinux_2_17_armv7l.manylinux_2_31_armv7l.whl", hash = "sha256:74bb723680f9f7a6234dcf67aea57e708ec1fbdf5699fb91dfd6f511b0a320ef", size = 144558, upload-time = "2025-10-14T04:40:55.677Z" },
@@ -74,6 +42,38 @@ wheels = [
     { url = "https://pypi.tuna.tsinghua.edu.cn/packages/a8/ef/89297262b8092b312d29cdb2517cb1237e51db8ecef2e9af5edbe7b683b1/charset_normalizer-3.4.4-cp312-cp312-win32.whl", hash = "sha256:5833d2c39d8896e4e19b689ffc198f08ea58116bee26dea51e362ecc7cd3ed26", size = 99694, upload-time = "2025-10-14T04:41:09.23Z" },
     { url = "https://pypi.tuna.tsinghua.edu.cn/packages/3d/2d/1e5ed9dd3b3803994c155cd9aacb60c82c331bad84daf75bcb9c91b3295e/charset_normalizer-3.4.4-cp312-cp312-win_amd64.whl", hash = "sha256:a79cfe37875f822425b89a82333404539ae63dbdddf97f84dcbc3d339aae9525", size = 107131, upload-time = "2025-10-14T04:41:10.467Z" },
     { url = "https://pypi.tuna.tsinghua.edu.cn/packages/d0/d9/0ed4c7098a861482a7b6a95603edce4c0d9db2311af23da1fb2b75ec26fc/charset_normalizer-3.4.4-cp312-cp312-win_arm64.whl", hash = "sha256:376bec83a63b8021bb5c8ea75e21c4ccb86e7e45ca4eb81146091b56599b80c3", size = 100390, upload-time = "2025-10-14T04:41:11.915Z" },
+    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/97/45/4b3a1239bbacd321068ea6e7ac28875b03ab8bc0aa0966452db17cd36714/charset_normalizer-3.4.4-cp313-cp313-macosx_10_13_universal2.whl", hash = "sha256:e1f185f86a6f3403aa2420e815904c67b2f9ebc443f045edd0de921108345794", size = 208091, upload-time = "2025-10-14T04:41:13.346Z" },
+    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/7d/62/73a6d7450829655a35bb88a88fca7d736f9882a27eacdca2c6d505b57e2e/charset_normalizer-3.4.4-cp313-cp313-manylinux2014_aarch64.manylinux_2_17_aarch64.manylinux_2_28_aarch64.whl", hash = "sha256:6b39f987ae8ccdf0d2642338faf2abb1862340facc796048b604ef14919e55ed", size = 147936, upload-time = "2025-10-14T04:41:14.461Z" },
+    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/89/c5/adb8c8b3d6625bef6d88b251bbb0d95f8205831b987631ab0c8bb5d937c2/charset_normalizer-3.4.4-cp313-cp313-manylinux2014_armv7l.manylinux_2_17_armv7l.manylinux_2_31_armv7l.whl", hash = "sha256:3162d5d8ce1bb98dd51af660f2121c55d0fa541b46dff7bb9b9f86ea1d87de72", size = 144180, upload-time = "2025-10-14T04:41:15.588Z" },
+    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/91/ed/9706e4070682d1cc219050b6048bfd293ccf67b3d4f5a4f39207453d4b99/charset_normalizer-3.4.4-cp313-cp313-manylinux2014_ppc64le.manylinux_2_17_ppc64le.manylinux_2_28_ppc64le.whl", hash = "sha256:81d5eb2a312700f4ecaa977a8235b634ce853200e828fbadf3a9c50bab278328", size = 161346, upload-time = "2025-10-14T04:41:16.738Z" },
+    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/d5/0d/031f0d95e4972901a2f6f09ef055751805ff541511dc1252ba3ca1f80cf5/charset_normalizer-3.4.4-cp313-cp313-manylinux2014_s390x.manylinux_2_17_s390x.manylinux_2_28_s390x.whl", hash = "sha256:5bd2293095d766545ec1a8f612559f6b40abc0eb18bb2f5d1171872d34036ede", size = 158874, upload-time = "2025-10-14T04:41:17.923Z" },
+    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/f5/83/6ab5883f57c9c801ce5e5677242328aa45592be8a00644310a008d04f922/charset_normalizer-3.4.4-cp313-cp313-manylinux2014_x86_64.manylinux_2_17_x86_64.manylinux_2_28_x86_64.whl", hash = "sha256:a8a8b89589086a25749f471e6a900d3f662d1d3b6e2e59dcecf787b1cc3a1894", size = 153076, upload-time = "2025-10-14T04:41:19.106Z" },
+    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/75/1e/5ff781ddf5260e387d6419959ee89ef13878229732732ee73cdae01800f2/charset_normalizer-3.4.4-cp313-cp313-manylinux_2_31_riscv64.manylinux_2_39_riscv64.whl", hash = "sha256:bc7637e2f80d8530ee4a78e878bce464f70087ce73cf7c1caf142416923b98f1", size = 150601, upload-time = "2025-10-14T04:41:20.245Z" },
+    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/d7/57/71be810965493d3510a6ca79b90c19e48696fb1ff964da319334b12677f0/charset_normalizer-3.4.4-cp313-cp313-musllinux_1_2_aarch64.whl", hash = "sha256:f8bf04158c6b607d747e93949aa60618b61312fe647a6369f88ce2ff16043490", size = 150376, upload-time = "2025-10-14T04:41:21.398Z" },
+    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/e5/d5/c3d057a78c181d007014feb7e9f2e65905a6c4ef182c0ddf0de2924edd65/charset_normalizer-3.4.4-cp313-cp313-musllinux_1_2_armv7l.whl", hash = "sha256:554af85e960429cf30784dd47447d5125aaa3b99a6f0683589dbd27e2f45da44", size = 144825, upload-time = "2025-10-14T04:41:22.583Z" },
+    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/e6/8c/d0406294828d4976f275ffbe66f00266c4b3136b7506941d87c00cab5272/charset_normalizer-3.4.4-cp313-cp313-musllinux_1_2_ppc64le.whl", hash = "sha256:74018750915ee7ad843a774364e13a3db91682f26142baddf775342c3f5b1133", size = 162583, upload-time = "2025-10-14T04:41:23.754Z" },
+    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/d7/24/e2aa1f18c8f15c4c0e932d9287b8609dd30ad56dbe41d926bd846e22fb8d/charset_normalizer-3.4.4-cp313-cp313-musllinux_1_2_riscv64.whl", hash = "sha256:c0463276121fdee9c49b98908b3a89c39be45d86d1dbaa22957e38f6321d4ce3", size = 150366, upload-time = "2025-10-14T04:41:25.27Z" },
+    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/e4/5b/1e6160c7739aad1e2df054300cc618b06bf784a7a164b0f238360721ab86/charset_normalizer-3.4.4-cp313-cp313-musllinux_1_2_s390x.whl", hash = "sha256:362d61fd13843997c1c446760ef36f240cf81d3ebf74ac62652aebaf7838561e", size = 160300, upload-time = "2025-10-14T04:41:26.725Z" },
+    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/7a/10/f882167cd207fbdd743e55534d5d9620e095089d176d55cb22d5322f2afd/charset_normalizer-3.4.4-cp313-cp313-musllinux_1_2_x86_64.whl", hash = "sha256:9a26f18905b8dd5d685d6d07b0cdf98a79f3c7a918906af7cc143ea2e164c8bc", size = 154465, upload-time = "2025-10-14T04:41:28.322Z" },
+    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/89/66/c7a9e1b7429be72123441bfdbaf2bc13faab3f90b933f664db506dea5915/charset_normalizer-3.4.4-cp313-cp313-win32.whl", hash = "sha256:9b35f4c90079ff2e2edc5b26c0c77925e5d2d255c42c74fdb70fb49b172726ac", size = 99404, upload-time = "2025-10-14T04:41:29.95Z" },
+    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/c4/26/b9924fa27db384bdcd97ab83b4f0a8058d96ad9626ead570674d5e737d90/charset_normalizer-3.4.4-cp313-cp313-win_amd64.whl", hash = "sha256:b435cba5f4f750aa6c0a0d92c541fb79f69a387c91e61f1795227e4ed9cece14", size = 107092, upload-time = "2025-10-14T04:41:31.188Z" },
+    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/af/8f/3ed4bfa0c0c72a7ca17f0380cd9e4dd842b09f664e780c13cff1dcf2ef1b/charset_normalizer-3.4.4-cp313-cp313-win_arm64.whl", hash = "sha256:542d2cee80be6f80247095cc36c418f7bddd14f4a6de45af91dfad36d817bba2", size = 100408, upload-time = "2025-10-14T04:41:32.624Z" },
+    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/2a/35/7051599bd493e62411d6ede36fd5af83a38f37c4767b92884df7301db25d/charset_normalizer-3.4.4-cp314-cp314-macosx_10_13_universal2.whl", hash = "sha256:da3326d9e65ef63a817ecbcc0df6e94463713b754fe293eaa03da99befb9a5bd", size = 207746, upload-time = "2025-10-14T04:41:33.773Z" },
+    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/10/9a/97c8d48ef10d6cd4fcead2415523221624bf58bcf68a802721a6bc807c8f/charset_normalizer-3.4.4-cp314-cp314-manylinux2014_aarch64.manylinux_2_17_aarch64.manylinux_2_28_aarch64.whl", hash = "sha256:8af65f14dc14a79b924524b1e7fffe304517b2bff5a58bf64f30b98bbc5079eb", size = 147889, upload-time = "2025-10-14T04:41:34.897Z" },
+    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/10/bf/979224a919a1b606c82bd2c5fa49b5c6d5727aa47b4312bb27b1734f53cd/charset_normalizer-3.4.4-cp314-cp314-manylinux2014_armv7l.manylinux_2_17_armv7l.manylinux_2_31_armv7l.whl", hash = "sha256:74664978bb272435107de04e36db5a9735e78232b85b77d45cfb38f758efd33e", size = 143641, upload-time = "2025-10-14T04:41:36.116Z" },
+    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/ba/33/0ad65587441fc730dc7bd90e9716b30b4702dc7b617e6ba4997dc8651495/charset_normalizer-3.4.4-cp314-cp314-manylinux2014_ppc64le.manylinux_2_17_ppc64le.manylinux_2_28_ppc64le.whl", hash = "sha256:752944c7ffbfdd10c074dc58ec2d5a8a4cd9493b314d367c14d24c17684ddd14", size = 160779, upload-time = "2025-10-14T04:41:37.229Z" },
+    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/67/ed/331d6b249259ee71ddea93f6f2f0a56cfebd46938bde6fcc6f7b9a3d0e09/charset_normalizer-3.4.4-cp314-cp314-manylinux2014_s390x.manylinux_2_17_s390x.manylinux_2_28_s390x.whl", hash = "sha256:d1f13550535ad8cff21b8d757a3257963e951d96e20ec82ab44bc64aeb62a191", size = 159035, upload-time = "2025-10-14T04:41:38.368Z" },
+    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/67/ff/f6b948ca32e4f2a4576aa129d8bed61f2e0543bf9f5f2b7fc3758ed005c9/charset_normalizer-3.4.4-cp314-cp314-manylinux2014_x86_64.manylinux_2_17_x86_64.manylinux_2_28_x86_64.whl", hash = "sha256:ecaae4149d99b1c9e7b88bb03e3221956f68fd6d50be2ef061b2381b61d20838", size = 152542, upload-time = "2025-10-14T04:41:39.862Z" },
+    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/16/85/276033dcbcc369eb176594de22728541a925b2632f9716428c851b149e83/charset_normalizer-3.4.4-cp314-cp314-manylinux_2_31_riscv64.manylinux_2_39_riscv64.whl", hash = "sha256:cb6254dc36b47a990e59e1068afacdcd02958bdcce30bb50cc1700a8b9d624a6", size = 149524, upload-time = "2025-10-14T04:41:41.319Z" },
+    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/9e/f2/6a2a1f722b6aba37050e626530a46a68f74e63683947a8acff92569f979a/charset_normalizer-3.4.4-cp314-cp314-musllinux_1_2_aarch64.whl", hash = "sha256:c8ae8a0f02f57a6e61203a31428fa1d677cbe50c93622b4149d5c0f319c1d19e", size = 150395, upload-time = "2025-10-14T04:41:42.539Z" },
+    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/60/bb/2186cb2f2bbaea6338cad15ce23a67f9b0672929744381e28b0592676824/charset_normalizer-3.4.4-cp314-cp314-musllinux_1_2_armv7l.whl", hash = "sha256:47cc91b2f4dd2833fddaedd2893006b0106129d4b94fdb6af1f4ce5a9965577c", size = 143680, upload-time = "2025-10-14T04:41:43.661Z" },
+    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/7d/a5/bf6f13b772fbb2a90360eb620d52ed8f796f3c5caee8398c3b2eb7b1c60d/charset_normalizer-3.4.4-cp314-cp314-musllinux_1_2_ppc64le.whl", hash = "sha256:82004af6c302b5d3ab2cfc4cc5f29db16123b1a8417f2e25f9066f91d4411090", size = 162045, upload-time = "2025-10-14T04:41:44.821Z" },
+    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/df/c5/d1be898bf0dc3ef9030c3825e5d3b83f2c528d207d246cbabe245966808d/charset_normalizer-3.4.4-cp314-cp314-musllinux_1_2_riscv64.whl", hash = "sha256:2b7d8f6c26245217bd2ad053761201e9f9680f8ce52f0fcd8d0755aeae5b2152", size = 149687, upload-time = "2025-10-14T04:41:46.442Z" },
+    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/a5/42/90c1f7b9341eef50c8a1cb3f098ac43b0508413f33affd762855f67a410e/charset_normalizer-3.4.4-cp314-cp314-musllinux_1_2_s390x.whl", hash = "sha256:799a7a5e4fb2d5898c60b640fd4981d6a25f1c11790935a44ce38c54e985f828", size = 160014, upload-time = "2025-10-14T04:41:47.631Z" },
+    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/76/be/4d3ee471e8145d12795ab655ece37baed0929462a86e72372fd25859047c/charset_normalizer-3.4.4-cp314-cp314-musllinux_1_2_x86_64.whl", hash = "sha256:99ae2cffebb06e6c22bdc25801d7b30f503cc87dbd283479e7b606f70aff57ec", size = 154044, upload-time = "2025-10-14T04:41:48.81Z" },
+    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/b0/6f/8f7af07237c34a1defe7defc565a9bc1807762f672c0fde711a4b22bf9c0/charset_normalizer-3.4.4-cp314-cp314-win32.whl", hash = "sha256:f9d332f8c2a2fcbffe1378594431458ddbef721c1769d78e2cbc06280d8155f9", size = 99940, upload-time = "2025-10-14T04:41:49.946Z" },
+    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/4b/51/8ade005e5ca5b0d80fb4aff72a3775b325bdc3d27408c8113811a7cbe640/charset_normalizer-3.4.4-cp314-cp314-win_amd64.whl", hash = "sha256:8a6562c3700cce886c5be75ade4a5db4214fda19fede41d9792d100288d8f94c", size = 107104, upload-time = "2025-10-14T04:41:51.051Z" },
+    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/da/5f/6b8f83a55bb8278772c5ae54a577f3099025f9ade59d0136ac24a0df4bde/charset_normalizer-3.4.4-cp314-cp314-win_arm64.whl", hash = "sha256:de00632ca48df9daf77a2c65a484531649261ec9f25489917f09e455cb09ddb2", size = 100743, upload-time = "2025-10-14T04:41:52.122Z" },
     { url = "https://pypi.tuna.tsinghua.edu.cn/packages/0a/4c/925909008ed5a988ccbb72dcc897407e5d6d3bd72410d69e051fc0c14647/charset_normalizer-3.4.4-py3-none-any.whl", hash = "sha256:7a32c560861a02ff789ad905a2fe94e3f840803362c84fecf1851cb4cf3dc37f", size = 53402, upload-time = "2025-10-14T04:42:31.76Z" },
 ]
 
@@ -86,18 +86,6 @@ wheels = [
     { url = "https://pypi.tuna.tsinghua.edu.cn/packages/d1/d6/3965ed04c63042e047cb6a3e6ed1a63a35087b6a609aa3a15ed8ac56c221/colorama-0.4.6-py2.py3-none-any.whl", hash = "sha256:4f1d9991f5acc0ca119f9d443620b77f9d6b33703e51011c16baf57afb285fc6", size = 25335, upload-time = "2022-10-25T02:36:20.889Z" },
 ]
 
-[[package]]
-name = "exceptiongroup"
-version = "1.3.1"
-source = { registry = "https://pypi.tuna.tsinghua.edu.cn/simple" }
-dependencies = [
-    { name = "typing-extensions" },
-]
-sdist = { url = "https://pypi.tuna.tsinghua.edu.cn/packages/50/79/66800aadf48771f6b62f7eb014e352e5d06856655206165d775e675a02c9/exceptiongroup-1.3.1.tar.gz", hash = "sha256:8b412432c6055b0b7d14c310000ae93352ed6754f70fa8f7c34141f91c4e3219", size = 30371, upload-time = "2025-11-21T23:01:54.787Z" }
-wheels = [
-    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/8a/0e/97c33bf5009bdbac74fd2beace167cab3f978feb69cc36f1ef79360d6c4e/exceptiongroup-1.3.1-py3-none-any.whl", hash = "sha256:a7a39a3bd276781e98394987d3a5701d0c4edffb633bb7a5144577f82c773598", size = 16740, upload-time = "2025-11-21T23:01:53.443Z" },
-]
-
 [[package]]
 name = "idna"
 version = "3.11"
@@ -149,6 +137,17 @@ version = "3.23.0"
 source = { registry = "https://pypi.tuna.tsinghua.edu.cn/simple" }
 sdist = { url = "https://pypi.tuna.tsinghua.edu.cn/packages/c9/85/e24bf90972a30b0fcd16c73009add1d7d7cd9140c2498a68252028899e41/pycryptodomex-3.23.0.tar.gz", hash = "sha256:71909758f010c82bc99b0abf4ea12012c98962fbf0583c2164f8b84533c2e4da", size = 4922157, upload-time = "2025-05-17T17:23:41.434Z" }
 wheels = [
+    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/2e/00/10edb04777069a42490a38c137099d4b17ba6e36a4e6e28bdc7470e9e853/pycryptodomex-3.23.0-cp313-cp313t-macosx_10_13_universal2.whl", hash = "sha256:7b37e08e3871efe2187bc1fd9320cc81d87caf19816c648f24443483005ff886", size = 2498764, upload-time = "2025-05-17T17:22:21.453Z" },
+    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/6b/3f/2872a9c2d3a27eac094f9ceaa5a8a483b774ae69018040ea3240d5b11154/pycryptodomex-3.23.0-cp313-cp313t-macosx_10_13_x86_64.whl", hash = "sha256:91979028227543010d7b2ba2471cf1d1e398b3f183cb105ac584df0c36dac28d", size = 1643012, upload-time = "2025-05-17T17:22:23.702Z" },
+    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/70/af/774c2e2b4f6570fbf6a4972161adbb183aeeaa1863bde31e8706f123bf92/pycryptodomex-3.23.0-cp313-cp313t-manylinux_2_17_aarch64.manylinux2014_aarch64.whl", hash = "sha256:6b8962204c47464d5c1c4038abeadd4514a133b28748bcd9fa5b6d62e3cec6fa", size = 2187643, upload-time = "2025-05-17T17:22:26.37Z" },
+    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/de/a3/71065b24cb889d537954cedc3ae5466af00a2cabcff8e29b73be047e9a19/pycryptodomex-3.23.0-cp313-cp313t-manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:a33986a0066860f7fcf7c7bd2bc804fa90e434183645595ae7b33d01f3c91ed8", size = 2273762, upload-time = "2025-05-17T17:22:28.313Z" },
+    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/c9/0b/ff6f43b7fbef4d302c8b981fe58467b8871902cdc3eb28896b52421422cc/pycryptodomex-3.23.0-cp313-cp313t-manylinux_2_5_i686.manylinux1_i686.manylinux_2_17_i686.manylinux2014_i686.whl", hash = "sha256:c7947ab8d589e3178da3d7cdeabe14f841b391e17046954f2fbcd941705762b5", size = 2313012, upload-time = "2025-05-17T17:22:30.57Z" },
+    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/02/de/9d4772c0506ab6da10b41159493657105d3f8bb5c53615d19452afc6b315/pycryptodomex-3.23.0-cp313-cp313t-musllinux_1_2_aarch64.whl", hash = "sha256:c25e30a20e1b426e1f0fa00131c516f16e474204eee1139d1603e132acffc314", size = 2186856, upload-time = "2025-05-17T17:22:32.819Z" },
+    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/28/ad/8b30efcd6341707a234e5eba5493700a17852ca1ac7a75daa7945fcf6427/pycryptodomex-3.23.0-cp313-cp313t-musllinux_1_2_i686.whl", hash = "sha256:da4fa650cef02db88c2b98acc5434461e027dce0ae8c22dd5a69013eaf510006", size = 2347523, upload-time = "2025-05-17T17:22:35.386Z" },
+    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/0f/02/16868e9f655b7670dbb0ac4f2844145cbc42251f916fc35c414ad2359849/pycryptodomex-3.23.0-cp313-cp313t-musllinux_1_2_x86_64.whl", hash = "sha256:58b851b9effd0d072d4ca2e4542bf2a4abcf13c82a29fd2c93ce27ee2a2e9462", size = 2272825, upload-time = "2025-05-17T17:22:37.632Z" },
+    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/ca/18/4ca89ac737230b52ac8ffaca42f9c6f1fd07c81a6cd821e91af79db60632/pycryptodomex-3.23.0-cp313-cp313t-win32.whl", hash = "sha256:a9d446e844f08299236780f2efa9898c818fe7e02f17263866b8550c7d5fb328", size = 1772078, upload-time = "2025-05-17T17:22:40Z" },
+    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/73/34/13e01c322db027682e00986873eca803f11c56ade9ba5bbf3225841ea2d4/pycryptodomex-3.23.0-cp313-cp313t-win_amd64.whl", hash = "sha256:bc65bdd9fc8de7a35a74cab1c898cab391a4add33a8fe740bda00f5976ca4708", size = 1803656, upload-time = "2025-05-17T17:22:42.139Z" },
+    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/54/68/9504c8796b1805d58f4425002bcca20f12880e6fa4dc2fc9a668705c7a08/pycryptodomex-3.23.0-cp313-cp313t-win_arm64.whl", hash = "sha256:c885da45e70139464f082018ac527fdaad26f1657a99ee13eecdce0f0ca24ab4", size = 1707172, upload-time = "2025-05-17T17:22:44.704Z" },
     { url = "https://pypi.tuna.tsinghua.edu.cn/packages/dd/9c/1a8f35daa39784ed8adf93a694e7e5dc15c23c741bbda06e1d45f8979e9e/pycryptodomex-3.23.0-cp37-abi3-macosx_10_9_universal2.whl", hash = "sha256:06698f957fe1ab229a99ba2defeeae1c09af185baa909a31a5d1f9d42b1aaed6", size = 2499240, upload-time = "2025-05-17T17:22:46.953Z" },
     { url = "https://pypi.tuna.tsinghua.edu.cn/packages/7a/62/f5221a191a97157d240cf6643747558759126c76ee92f29a3f4aee3197a5/pycryptodomex-3.23.0-cp37-abi3-macosx_10_9_x86_64.whl", hash = "sha256:b2c2537863eccef2d41061e82a881dcabb04944c5c06c5aa7110b577cc487545", size = 1644042, upload-time = "2025-05-17T17:22:49.098Z" },
     { url = "https://pypi.tuna.tsinghua.edu.cn/packages/8c/fd/5a054543c8988d4ed7b612721d7e78a4b9bf36bc3c5ad45ef45c22d0060e/pycryptodomex-3.23.0-cp37-abi3-manylinux_2_17_aarch64.manylinux2014_aarch64.whl", hash = "sha256:43c446e2ba8df8889e0e16f02211c25b4934898384c1ec1ec04d7889c0333587", size = 2186227, upload-time = "2025-05-17T17:22:51.139Z" },
@@ -160,11 +159,6 @@ wheels = [
     { url = "https://pypi.tuna.tsinghua.edu.cn/packages/8d/67/09ee8500dd22614af5fbaa51a4aee6e342b5fa8aecf0a6cb9cbf52fa6d45/pycryptodomex-3.23.0-cp37-abi3-win32.whl", hash = "sha256:189afbc87f0b9f158386bf051f720e20fa6145975f1e76369303d0f31d1a8d7c", size = 1771969, upload-time = "2025-05-17T17:23:07.115Z" },
     { url = "https://pypi.tuna.tsinghua.edu.cn/packages/69/96/11f36f71a865dd6df03716d33bd07a67e9d20f6b8d39820470b766af323c/pycryptodomex-3.23.0-cp37-abi3-win_amd64.whl", hash = "sha256:52e5ca58c3a0b0bd5e100a9fbc8015059b05cffc6c66ce9d98b4b45e023443b9", size = 1803124, upload-time = "2025-05-17T17:23:09.267Z" },
     { url = "https://pypi.tuna.tsinghua.edu.cn/packages/f9/93/45c1cdcbeb182ccd2e144c693eaa097763b08b38cded279f0053ed53c553/pycryptodomex-3.23.0-cp37-abi3-win_arm64.whl", hash = "sha256:02d87b80778c171445d67e23d1caef279bf4b25c3597050ccd2e13970b57fd51", size = 1707161, upload-time = "2025-05-17T17:23:11.414Z" },
-    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/f3/b8/3e76d948c3c4ac71335bbe75dac53e154b40b0f8f1f022dfa295257a0c96/pycryptodomex-3.23.0-pp310-pypy310_pp73-macosx_10_15_x86_64.whl", hash = "sha256:ebfff755c360d674306e5891c564a274a47953562b42fb74a5c25b8fc1fb1cb5", size = 1627695, upload-time = "2025-05-17T17:23:17.38Z" },
-    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/6a/cf/80f4297a4820dfdfd1c88cf6c4666a200f204b3488103d027b5edd9176ec/pycryptodomex-3.23.0-pp310-pypy310_pp73-manylinux_2_17_aarch64.manylinux2014_aarch64.whl", hash = "sha256:eca54f4bb349d45afc17e3011ed4264ef1cc9e266699874cdd1349c504e64798", size = 1675772, upload-time = "2025-05-17T17:23:19.202Z" },
-    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/d1/42/1e969ee0ad19fe3134b0e1b856c39bd0b70d47a4d0e81c2a8b05727394c9/pycryptodomex-3.23.0-pp310-pypy310_pp73-manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:4f2596e643d4365e14d0879dc5aafe6355616c61c2176009270f3048f6d9a61f", size = 1668083, upload-time = "2025-05-17T17:23:21.867Z" },
-    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/6e/c3/1de4f7631fea8a992a44ba632aa40e0008764c0fb9bf2854b0acf78c2cf2/pycryptodomex-3.23.0-pp310-pypy310_pp73-manylinux_2_5_i686.manylinux1_i686.manylinux_2_17_i686.manylinux2014_i686.whl", hash = "sha256:fdfac7cda115bca3a5abb2f9e43bc2fb66c2b65ab074913643803ca7083a79ea", size = 1706056, upload-time = "2025-05-17T17:23:24.031Z" },
-    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/f2/5f/af7da8e6f1e42b52f44a24d08b8e4c726207434e2593732d39e7af5e7256/pycryptodomex-3.23.0-pp310-pypy310_pp73-win_amd64.whl", hash = "sha256:14c37aaece158d0ace436f76a7bb19093db3b4deade9797abfc39ec6cd6cc2fe", size = 1806478, upload-time = "2025-05-17T17:23:26.066Z" },
 ]
 
 [[package]]
@@ -182,12 +176,10 @@ version = "9.0.1"
 source = { registry = "https://pypi.tuna.tsinghua.edu.cn/simple" }
 dependencies = [
     { name = "colorama", marker = "sys_platform == 'win32'" },
-    { name = "exceptiongroup", marker = "python_full_version < '3.11'" },
     { name = "iniconfig" },
     { name = "packaging" },
     { name = "pluggy" },
     { name = "pygments" },
-    { name = "tomli", marker = "python_full_version < '3.11'" },
 ]
 sdist = { url = "https://pypi.tuna.tsinghua.edu.cn/packages/07/56/f013048ac4bc4c1d9be45afd4ab209ea62822fb1598f40687e6bf45dcea4/pytest-9.0.1.tar.gz", hash = "sha256:3e9c069ea73583e255c3b21cf46b8d3c56f6e3a1a8f6da94ccb0fcf57b9d73c8", size = 1564125, upload-time = "2025-11-12T13:05:09.333Z" }
 wheels = [
@@ -196,7 +188,7 @@ wheels = [
 
 [[package]]
 name = "ragflow-cli"
-version = "0.24.0"
+version = "0.25.0"
 source = { virtual = "." }
 dependencies = [
     { name = "beartype" },
@@ -254,45 +246,11 @@ wheels = [
     { url = "https://pypi.tuna.tsinghua.edu.cn/packages/3f/51/d4db610ef29373b879047326cbf6fa98b6c1969d6f6dc423279de2b1be2c/requests_toolbelt-1.0.0-py2.py3-none-any.whl", hash = "sha256:cccfdd665f0a24fcf4726e690f65639d272bb0637b9b92dfd91a5568ccf6bd06", size = 54481, upload-time = "2023-05-01T04:11:28.427Z" },
 ]
 
-[[package]]
-name = "tomli"
-version = "2.3.0"
-source = { registry = "https://pypi.tuna.tsinghua.edu.cn/simple" }
-sdist = { url = "https://pypi.tuna.tsinghua.edu.cn/packages/52/ed/3f73f72945444548f33eba9a87fc7a6e969915e7b1acc8260b30e1f76a2f/tomli-2.3.0.tar.gz", hash = "sha256:64be704a875d2a59753d80ee8a533c3fe183e3f06807ff7dc2232938ccb01549", size = 17392, upload-time = "2025-10-08T22:01:47.119Z" }
-wheels = [
-    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/b3/2e/299f62b401438d5fe1624119c723f5d877acc86a4c2492da405626665f12/tomli-2.3.0-cp311-cp311-macosx_10_9_x86_64.whl", hash = "sha256:88bd15eb972f3664f5ed4b57c1634a97153b4bac4479dcb6a495f41921eb7f45", size = 153236, upload-time = "2025-10-08T22:01:00.137Z" },
-    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/86/7f/d8fffe6a7aefdb61bced88fcb5e280cfd71e08939da5894161bd71bea022/tomli-2.3.0-cp311-cp311-macosx_11_0_arm64.whl", hash = "sha256:883b1c0d6398a6a9d29b508c331fa56adbcdff647f6ace4dfca0f50e90dfd0ba", size = 148084, upload-time = "2025-10-08T22:01:01.63Z" },
-    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/47/5c/24935fb6a2ee63e86d80e4d3b58b222dafaf438c416752c8b58537c8b89a/tomli-2.3.0-cp311-cp311-manylinux2014_aarch64.manylinux_2_17_aarch64.manylinux_2_28_aarch64.whl", hash = "sha256:d1381caf13ab9f300e30dd8feadb3de072aeb86f1d34a8569453ff32a7dea4bf", size = 234832, upload-time = "2025-10-08T22:01:02.543Z" },
-    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/89/da/75dfd804fc11e6612846758a23f13271b76d577e299592b4371a4ca4cd09/tomli-2.3.0-cp311-cp311-manylinux2014_x86_64.manylinux_2_17_x86_64.manylinux_2_28_x86_64.whl", hash = "sha256:a0e285d2649b78c0d9027570d4da3425bdb49830a6156121360b3f8511ea3441", size = 242052, upload-time = "2025-10-08T22:01:03.836Z" },
-    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/70/8c/f48ac899f7b3ca7eb13af73bacbc93aec37f9c954df3c08ad96991c8c373/tomli-2.3.0-cp311-cp311-musllinux_1_2_aarch64.whl", hash = "sha256:0a154a9ae14bfcf5d8917a59b51ffd5a3ac1fd149b71b47a3a104ca4edcfa845", size = 239555, upload-time = "2025-10-08T22:01:04.834Z" },
-    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/ba/28/72f8afd73f1d0e7829bfc093f4cb98ce0a40ffc0cc997009ee1ed94ba705/tomli-2.3.0-cp311-cp311-musllinux_1_2_x86_64.whl", hash = "sha256:74bf8464ff93e413514fefd2be591c3b0b23231a77f901db1eb30d6f712fc42c", size = 245128, upload-time = "2025-10-08T22:01:05.84Z" },
-    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/b6/eb/a7679c8ac85208706d27436e8d421dfa39d4c914dcf5fa8083a9305f58d9/tomli-2.3.0-cp311-cp311-win32.whl", hash = "sha256:00b5f5d95bbfc7d12f91ad8c593a1659b6387b43f054104cda404be6bda62456", size = 96445, upload-time = "2025-10-08T22:01:06.896Z" },
-    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/0a/fe/3d3420c4cb1ad9cb462fb52967080575f15898da97e21cb6f1361d505383/tomli-2.3.0-cp311-cp311-win_amd64.whl", hash = "sha256:4dc4ce8483a5d429ab602f111a93a6ab1ed425eae3122032db7e9acf449451be", size = 107165, upload-time = "2025-10-08T22:01:08.107Z" },
-    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/ff/b7/40f36368fcabc518bb11c8f06379a0fd631985046c038aca08c6d6a43c6e/tomli-2.3.0-cp312-cp312-macosx_10_13_x86_64.whl", hash = "sha256:d7d86942e56ded512a594786a5ba0a5e521d02529b3826e7761a05138341a2ac", size = 154891, upload-time = "2025-10-08T22:01:09.082Z" },
-    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/f9/3f/d9dd692199e3b3aab2e4e4dd948abd0f790d9ded8cd10cbaae276a898434/tomli-2.3.0-cp312-cp312-macosx_11_0_arm64.whl", hash = "sha256:73ee0b47d4dad1c5e996e3cd33b8a76a50167ae5f96a2607cbe8cc773506ab22", size = 148796, upload-time = "2025-10-08T22:01:10.266Z" },
-    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/60/83/59bff4996c2cf9f9387a0f5a3394629c7efa5ef16142076a23a90f1955fa/tomli-2.3.0-cp312-cp312-manylinux2014_aarch64.manylinux_2_17_aarch64.manylinux_2_28_aarch64.whl", hash = "sha256:792262b94d5d0a466afb5bc63c7daa9d75520110971ee269152083270998316f", size = 242121, upload-time = "2025-10-08T22:01:11.332Z" },
-    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/45/e5/7c5119ff39de8693d6baab6c0b6dcb556d192c165596e9fc231ea1052041/tomli-2.3.0-cp312-cp312-manylinux2014_x86_64.manylinux_2_17_x86_64.manylinux_2_28_x86_64.whl", hash = "sha256:4f195fe57ecceac95a66a75ac24d9d5fbc98ef0962e09b2eddec5d39375aae52", size = 250070, upload-time = "2025-10-08T22:01:12.498Z" },
-    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/45/12/ad5126d3a278f27e6701abde51d342aa78d06e27ce2bb596a01f7709a5a2/tomli-2.3.0-cp312-cp312-musllinux_1_2_aarch64.whl", hash = "sha256:e31d432427dcbf4d86958c184b9bfd1e96b5b71f8eb17e6d02531f434fd335b8", size = 245859, upload-time = "2025-10-08T22:01:13.551Z" },
-    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/fb/a1/4d6865da6a71c603cfe6ad0e6556c73c76548557a8d658f9e3b142df245f/tomli-2.3.0-cp312-cp312-musllinux_1_2_x86_64.whl", hash = "sha256:7b0882799624980785240ab732537fcfc372601015c00f7fc367c55308c186f6", size = 250296, upload-time = "2025-10-08T22:01:14.614Z" },
-    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/a0/b7/a7a7042715d55c9ba6e8b196d65d2cb662578b4d8cd17d882d45322b0d78/tomli-2.3.0-cp312-cp312-win32.whl", hash = "sha256:ff72b71b5d10d22ecb084d345fc26f42b5143c5533db5e2eaba7d2d335358876", size = 97124, upload-time = "2025-10-08T22:01:15.629Z" },
-    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/06/1e/f22f100db15a68b520664eb3328fb0ae4e90530887928558112c8d1f4515/tomli-2.3.0-cp312-cp312-win_amd64.whl", hash = "sha256:1cb4ed918939151a03f33d4242ccd0aa5f11b3547d0cf30f7c74a408a5b99878", size = 107698, upload-time = "2025-10-08T22:01:16.51Z" },
-    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/77/b8/0135fadc89e73be292b473cb820b4f5a08197779206b33191e801feeae40/tomli-2.3.0-py3-none-any.whl", hash = "sha256:e95b1af3c5b07d9e643909b5abbec77cd9f1217e6d0bca72b0234736b9fb1f1b", size = 14408, upload-time = "2025-10-08T22:01:46.04Z" },
-]
-
-[[package]]
-name = "typing-extensions"
-version = "4.15.0"
-source = { registry = "https://pypi.tuna.tsinghua.edu.cn/simple" }
-sdist = { url = "https://pypi.tuna.tsinghua.edu.cn/packages/72/94/1a15dd82efb362ac84269196e94cf00f187f7ed21c242792a923cdb1c61f/typing_extensions-4.15.0.tar.gz", hash = "sha256:0cea48d173cc12fa28ecabc3b837ea3cf6f38c6d1136f85cbaaf598984861466", size = 109391, upload-time = "2025-08-25T13:49:26.313Z" }
-wheels = [
-    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/18/67/36e9267722cc04a6b9f15c7f3441c2363321a3ea07da7ae0c0707beb2a9c/typing_extensions-4.15.0-py3-none-any.whl", hash = "sha256:f0fa19c6845758ab08074a0cfa8b7aecb71c999ca73d62883bc25cc018c4e548", size = 44614, upload-time = "2025-08-25T13:49:24.86Z" },
-]
-
 [[package]]
 name = "urllib3"
-version = "2.5.0"
+version = "2.6.3"
 source = { registry = "https://pypi.tuna.tsinghua.edu.cn/simple" }
-sdist = { url = "https://pypi.tuna.tsinghua.edu.cn/packages/15/22/9ee70a2574a4f4599c47dd506532914ce044817c7752a79b6a51286319bc/urllib3-2.5.0.tar.gz", hash = "sha256:3fc47733c7e419d4bc3f6b3dc2b4f890bb743906a30d56ba4a5bfa4bbff92760", size = 393185, upload-time = "2025-06-18T14:07:41.644Z" }
+sdist = { url = "https://pypi.tuna.tsinghua.edu.cn/packages/c7/24/5f1b3bdffd70275f6661c76461e25f024d5a38a46f04aaca912426a2b1d3/urllib3-2.6.3.tar.gz", hash = "sha256:1b62b6884944a57dbe321509ab94fd4d3b307075e0c2eae991ac71ee15ad38ed", size = 435556, upload-time = "2026-01-07T16:24:43.925Z" }
 wheels = [
-    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/a7/c2/fe1e52489ae3122415c51f387e221dd0773709bad6c6cdaa599e8a2c5185/urllib3-2.5.0-py3-none-any.whl", hash = "sha256:e6b01673c0fa6a13e374b50871808eb3bf7046c4b125b216f6bf1cc604cff0dc", size = 129795, upload-time = "2025-06-18T14:07:40.39Z" },
+    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/39/08/aaaad47bc4e9dc8c725e68f9d04865dbcb2052843ff09c97b08904852d84/urllib3-2.6.3-py3-none-any.whl", hash = "sha256:bf272323e553dfb2e87d9bfd225ca7b0f467b919d7bbd355436d3fd37cb0acd4", size = 131584, upload-time = "2026-01-07T16:24:42.685Z" },
 ]

admin/server/admin_server.py

@@ -21,7 +21,6 @@ import os
 import signal
 import logging
 import threading
-import traceback
 import faulthandler
 
 from flask import Flask
@@ -58,7 +57,7 @@ if __name__ == '__main__':
         os.environ.get("MAX_CONTENT_LENGTH", 1024 * 1024 * 1024)
     )
     Session(app)
-    logging.info(f'RAGFlow version: {get_ragflow_version()}')
+    logging.info(f'RAGFlow admin version: {get_ragflow_version()}')
     show_configs()
     login_manager = LoginManager()
     login_manager.init_app(app)
@@ -75,10 +74,10 @@ if __name__ == '__main__':
             application=app,
             threaded=True,
             use_reloader=False,
-            use_debugger=True,
+            use_debugger=False,
         )
-    except Exception:
-        traceback.print_exc()
+    except Exception as e:
+        logging.exception(f"Unhandled exception: {e}")
         stop_event.set()
         time.sleep(1)
         os.kill(os.getpid(), signal.SIGKILL)

admin/server/auth.py

@@ -22,7 +22,6 @@ from datetime import datetime
 
 from flask import jsonify, request
 from flask_login import current_user, login_user
-from itsdangerous.url_safe import URLSafeTimedSerializer as Serializer
 
 from api.common.exceptions import AdminException, UserNotFoundError
 from api.common.base64 import encode_to_base64
@@ -40,18 +39,34 @@ from common import settings
 def setup_auth(login_manager):
     @login_manager.request_loader
     def load_user(web_request):
-        jwt = Serializer(secret_key=settings.SECRET_KEY)
+        # Authorization header contains JWT-encoded access token
+        # First decode JWT to get the UUID, then query database
+        from itsdangerous.url_safe import URLSafeTimedSerializer as Serializer
+        from common import settings
+
         authorization = web_request.headers.get("Authorization")
         if authorization:
             try:
-                access_token = str(jwt.loads(authorization))
+                # Strip "Bearer " prefix if present
+                jwt_token = authorization
+                if jwt_token.startswith("Bearer "):
+                    jwt_token = jwt_token[7:]
 
-                if not access_token or not access_token.strip():
-                    logging.warning("Authentication attempt with empty access token")
+                jwt_token = jwt_token.strip()
+                if not jwt_token:
+                    logging.warning("Authentication attempt with empty JWT token")
                     return None
 
-                # Access tokens should be UUIDs (32 hex characters)
-                if len(access_token.strip()) < 32:
+                # Decode JWT to get the UUID access_token
+                jwt = Serializer(secret_key=settings.SECRET_KEY)
+                access_token = str(jwt.loads(jwt_token))
+
+                if not access_token or not access_token.strip():
+                    logging.warning("Authentication attempt with empty access token after JWT decode")
+                    return None
+
+                # Access tokens stored in database are UUIDs (32 hex characters)
+                if len(access_token) < 32:
                     logging.warning(f"Authentication attempt with invalid token format: {len(access_token)} chars")
                     return None
 
@@ -110,7 +125,8 @@ def add_tenant_for_admin(user_info: dict, role: str):
         "embd_id": settings.EMBEDDING_MDL,
         "asr_id": settings.ASR_MDL,
         "parser_ids": settings.PARSERS,
-        "img2txt_id": settings.IMAGE2TEXT_MDL
+        "img2txt_id": settings.IMAGE2TEXT_MDL,
+        "rerank_id": settings.RERANK_MDL,
     }
     usr_tenant = {
         "tenant_id": user_info["id"],

admin/server/config.py

@@ -264,6 +264,19 @@ def load_configurations(config_path: str) -> list[BaseConfig]:
                                         db_name=database, detail_func_name="get_infinity_status")
                 configurations.append(config)
                 id_count += 1
+            case "minio_0":
+                name: str = 'minio_0'
+                url = v['host']
+                parts = url.split(':', 1)
+                host = parts[0]
+                port = int(parts[1])
+                user = v.get('user')
+                password = v.get('password')
+                config = MinioConfig(id=id_count, name=name, host=host, port=port, user=user, password=password,
+                                     service_type="file_store",
+                                     store_type="minio", detail_func_name="check_minio_alive")
+                configurations.append(config)
+                id_count += 1
             case "minio":
                 name: str = 'minio'
                 url = v['host']
@@ -310,6 +323,14 @@ def load_configurations(config_path: str) -> list[BaseConfig]:
                                             service_type="task_executor", detail_func_name="check_task_executor_alive")
                 configurations.append(config)
                 id_count += 1
+            case "rabbitmq":
+                name: str = 'rabbitmq'
+                host: str = v.get('host')
+                port: int = v.get('port')
+                config = RabbitMQConfig(id=id_count, name=name, host=host, port=port,
+                                        service_type="message_queue", mq_type="rabbitmq", detail_func_name="check_rabbitmq_alive")
+                configurations.append(config)
+                id_count += 1
             case _:
                 logging.warning(f"Unknown configuration key: {k}")
                 continue

admin/server/routes.py

@@ -30,13 +30,14 @@ from roles import RoleMgr
 from api.common.exceptions import AdminException
 from common.versions import get_ragflow_version
 from api.utils.api_utils import generate_confirmation_token
+from common.log_utils import get_log_levels, set_log_level
 
 admin_bp = Blueprint("admin", __name__, url_prefix="/api/v1/admin")
 
 
 @admin_bp.route("/ping", methods=["GET"])
 def ping():
-    return success_response("PONG")
+    return success_response(message="pong")
 
 
 @admin_bp.route("/login", methods=["POST"])
@@ -652,3 +653,39 @@ def test_sandbox_connection():
         return error_response(str(e), 400)
     except Exception as e:
         return error_response(str(e), 500)
+
+
+@admin_bp.route("/log_levels", methods=["GET"])
+@login_required
+@check_admin_auth
+def get_logger_levels():
+    """Get current log levels for all packages."""
+    try:
+        res = get_log_levels()
+        return success_response(res, "Get log levels", 0)
+    except Exception as e:
+        return error_response(str(e), 500)
+
+
+@admin_bp.route("/log_levels", methods=["PUT"])
+@login_required
+@check_admin_auth
+def set_logger_level():
+    """Set log level for a package."""
+    try:
+        data = request.get_json()
+        if not data or "pkg_name" not in data or "level" not in data:
+            return error_response("pkg_name and level are required", 400)
+
+        pkg_name = data["pkg_name"]
+        level = data["level"]
+        if not isinstance(pkg_name, str) or not isinstance(level, str):
+            return error_response("pkg_name and level must be strings", 400)
+
+        success = set_log_level(pkg_name, level)
+        if success:
+            return success_response({"pkg_name": pkg_name, "level": level}, "Log level updated successfully")
+        else:
+            return error_response(f"Invalid log level: {level}", 400)
+    except Exception as e:
+        return error_response(str(e), 500)

agent/canvas.py

@@ -15,6 +15,7 @@
 #
 import asyncio
 import base64
+import datetime
 import inspect
 import binascii
 import json
@@ -28,9 +29,11 @@ from typing import Any, Union, Tuple
 
 from agent.component import component_class
 from agent.component.base import ComponentBase
+from agent.dsl_migration import normalize_chunker_dsl
 from api.db.services.file_service import FileService
 from api.db.services.llm_service import LLMBundle
 from api.db.services.task_service import has_canceled
+from api.db.joint_services.tenant_model_service import get_tenant_default_model_by_type
 from common.constants import LLMType
 from common.misc_utils import get_uuid, hash_str2int
 from common.exceptions import TaskCanceledException
@@ -82,7 +85,8 @@ class Graph:
         self.path = []
         self.components = {}
         self.error = ""
-        self.dsl = json.loads(dsl)
+        # Accept legacy DSL on read, but keep the in-memory canvas in the latest schema.
+        self.dsl = normalize_chunker_dsl(json.loads(dsl))
         self._tenant_id = tenant_id
         self.task_id = task_id if task_id else get_uuid()
         self.custom_header = custom_header
@@ -286,7 +290,8 @@ class Canvas(Graph):
             "sys.user_id": tenant_id,
             "sys.conversation_turns": 0,
             "sys.files": [],
-            "sys.history": []
+            "sys.history": [],
+            "sys.date": datetime.datetime.now(datetime.timezone.utc).strftime("%Y-%m-%d %H:%M:%S")
         }
         self.variables = {}
         super().__init__(dsl, tenant_id, task_id, custom_header=custom_header)
@@ -299,13 +304,16 @@ class Canvas(Graph):
             self.globals = self.dsl["globals"]
             if "sys.history" not in self.globals:
                 self.globals["sys.history"] = []
+            if "sys.date" not in self.globals:
+                self.globals["sys.date"] = datetime.datetime.now(datetime.timezone.utc).strftime("%Y-%m-%d %H:%M:%S")
         else:
             self.globals = {
             "sys.query": "",
             "sys.user_id": "",
             "sys.conversation_turns": 0,
             "sys.files": [],
-            "sys.history": []
+            "sys.history": [],
+            "sys.date": datetime.datetime.now(datetime.timezone.utc).strftime("%Y-%m-%d %H:%M:%S")
         }
         if "variables" in self.dsl:
             self.variables = self.dsl["variables"]
@@ -367,6 +375,7 @@ class Canvas(Graph):
                     self.globals[k] = ""
 
     async def run(self, **kwargs):
+        self.globals["sys.date"] = datetime.datetime.now(datetime.timezone.utc).strftime("%Y-%m-%d %H:%M:%S")
         st = time.perf_counter()
         self._loop = asyncio.get_running_loop()
         self.message_id = get_uuid()
@@ -386,10 +395,16 @@ class Canvas(Graph):
                             continue
                         self.components[k]["obj"].set_output(kk, vv)
 
+        layout_recognize = None
+        for cpn in self.components.values():
+            if cpn["obj"].component_name.lower() == "begin":
+                layout_recognize = getattr(cpn["obj"]._param, "layout_recognize", None)
+                break
+
         for k in kwargs.keys():
             if k in ["query", "user_id", "files"] and kwargs[k]:
                 if k == "files":
-                    self.globals[f"sys.{k}"] = await self.get_files_async(kwargs[k])
+                    self.globals[f"sys.{k}"] = await self.get_files_async(kwargs[k], layout_recognize)
                 else:
                     self.globals[f"sys.{k}"] = kwargs[k]
         if not self.globals["sys.conversation_turns"] :
@@ -502,7 +517,8 @@ class Canvas(Graph):
                 cpn_obj = self.get_component_obj(self.path[i])
                 if cpn_obj.component_name.lower() == "message":
                     if cpn_obj.get_param("auto_play"):
-                        tts_mdl = LLMBundle(self._tenant_id, LLMType.TTS)
+                        tts_model_config = get_tenant_default_model_by_type(self._tenant_id, LLMType.TTS)
+                        tts_mdl = LLMBundle(self._tenant_id, tts_model_config)
                     if isinstance(cpn_obj.output("content"), partial):
                         _m = ""
                         buff_m = ""
@@ -547,18 +563,10 @@ class Canvas(Graph):
                             yield decorate("message", {"content": "", "audio_binary": self.tts(tts_mdl, buff_m)})
                             buff_m = ""
                         cpn_obj.set_output("content", _m)
-                        cite = re.search(r"\[ID:[ 0-9]+\]", _m)
                     else:
                         yield decorate("message", {"content": cpn_obj.output("content")})
-                        cite = re.search(r"\[ID:[ 0-9]+\]",  cpn_obj.output("content"))
 
-                    message_end = {}
-                    if cpn_obj.get_param("status"):
-                        message_end["status"] = cpn_obj.get_param("status")
-                    if isinstance(cpn_obj.output("attachment"), dict):
-                        message_end["attachment"] = cpn_obj.output("attachment")
-                    if cite:
-                        message_end["reference"] = self.get_reference()
+                    message_end = self._build_message_end(cpn_obj)
                     yield decorate("message_end", message_end)
 
                     while partials:
@@ -748,7 +756,7 @@ class Canvas(Graph):
     def get_component_input_elements(self, cpnnm):
         return self.components[cpnnm]["obj"].get_input_elements()
 
-    async def get_files_async(self, files: Union[None, list[dict]]) -> list[str]:
+    async def get_files_async(self, files: Union[None, list[dict]], layout_recognize: str = None) -> list[str]:
         if not files:
             return  []
         def image_to_base64(file):
@@ -756,7 +764,7 @@ class Canvas(Graph):
                                         base64.b64encode(FileService.get_blob(file["created_by"], file["id"])).decode("utf-8"))
         def parse_file(file):
             blob = FileService.get_blob(file["created_by"], file["id"])
-            return FileService.parse(file["name"], blob, True, file["created_by"])
+            return FileService.parse(file["name"], blob, True, file["created_by"], layout_recognize)
         loop = asyncio.get_running_loop()
         tasks = []
         for file in files:
@@ -766,15 +774,15 @@ class Canvas(Graph):
             tasks.append(loop.run_in_executor(self._thread_pool, parse_file, file))
         return await asyncio.gather(*tasks)
 
-    def get_files(self, files: Union[None, list[dict]]) -> list[str]:
+    def get_files(self, files: Union[None, list[dict]], layout_recognize: str = None) -> list[str]:
         """
         Synchronous wrapper for get_files_async, used by sync component invoke paths.
         """
         loop = getattr(self, "_loop", None)
         if loop and loop.is_running():
-            return asyncio.run_coroutine_threadsafe(self.get_files_async(files), loop).result()
+            return asyncio.run_coroutine_threadsafe(self.get_files_async(files, layout_recognize), loop).result()
 
-        return asyncio.run(self.get_files_async(files))
+        return asyncio.run(self.get_files_async(files, layout_recognize))
 
     def tool_use_callback(self, agent_id: str, func_name: str, params: dict, result: Any, elapsed_time=None):
         agent_ids = agent_id.split("-->")
@@ -820,6 +828,22 @@ class Canvas(Graph):
             return {"chunks": {}, "doc_aggs": {}}
         return self.retrieval[-1]
 
+    def _has_reference(self) -> bool:
+        ref = self.get_reference()
+        if not isinstance(ref, dict):
+            return False
+        return bool(ref.get("chunks") or ref.get("doc_aggs"))
+
+    def _build_message_end(self, cpn_obj) -> dict:
+        message_end = {}
+        if cpn_obj.get_param("status"):
+            message_end["status"] = cpn_obj.get_param("status")
+        if isinstance(cpn_obj.output("attachment"), dict):
+            message_end["attachment"] = cpn_obj.output("attachment")
+        if self._has_reference():
+            message_end["reference"] = self.get_reference()
+        return message_end
+
     def add_memory(self, user:str, assist:str, summ: str):
         self.memory.append((user, assist, summ))
 

agent/component/agent_with_tools.py

@@ -20,19 +20,20 @@ import os
 import re
 from copy import deepcopy
 from functools import partial
+from timeit import default_timer as timer
 from typing import Any
 
 import json_repair
-from timeit import default_timer as timer
-from agent.tools.base import LLMToolPluginCallSession, ToolParamBase, ToolBase, ToolMeta
+
+from agent.component.llm import LLM, LLMParam
+from agent.tools.base import LLMToolPluginCallSession, ToolBase, ToolMeta, ToolParamBase
+from api.db.joint_services.tenant_model_service import get_model_config_by_type_and_name
 from api.db.services.llm_service import LLMBundle
-from api.db.services.tenant_llm_service import TenantLLMService
 from api.db.services.mcp_server_service import MCPServerService
+from api.db.services.tenant_llm_service import TenantLLMService
 from common.connection_utils import timeout
-from rag.prompts.generator import next_step_async, COMPLETE_TASK, \
-    citation_prompt, kb_prompt, citation_plus, full_question, message_fit_in, structured_output_prompt
 from common.mcp_tool_call_conn import MCPToolCallSession, mcp_tool_metadata_to_openai_tool
-from agent.component.llm import LLMParam, LLM
+from rag.prompts.generator import citation_plus, citation_prompt, full_question, kb_prompt, message_fit_in, structured_output_prompt
 
 
 class AgentParam(LLMParam, ToolParamBase):
@@ -41,35 +42,25 @@ class AgentParam(LLMParam, ToolParamBase):
     """
 
     def __init__(self):
-        self.meta:ToolMeta = {
-                "name": "agent",
-                "description": "This is an agent for a specific task.",
-                "parameters": {
-                    "user_prompt": {
-                        "type": "string",
-                        "description": "This is the order you need to send to the agent.",
-                        "default": "",
-                        "required": True
-                    },
-                    "reasoning": {
-                        "type": "string",
-                        "description": (
-                            "Supervisor's reasoning for choosing the this agent. "
-                            "Explain why this agent is being invoked and what is expected of it."
-                        ),
-                        "required": True
-                    },
-                    "context": {
-                        "type": "string",
-                        "description": (
-                                "All relevant background information, prior facts, decisions, "
-                                "and state needed by the agent to solve the current query. "
-                                "Should be as detailed and self-contained as possible."
-                            ),
-                        "required": True
-                    },
-                }
-            }
+        self.meta: ToolMeta = {
+            "name": "agent",
+            "description": "This is an agent for a specific task.",
+            "parameters": {
+                "user_prompt": {"type": "string", "description": "This is the order you need to send to the agent.", "default": "", "required": True},
+                "reasoning": {
+                    "type": "string",
+                    "description": ("Supervisor's reasoning for choosing the this agent. Explain why this agent is being invoked and what is expected of it."),
+                    "required": True,
+                },
+                "context": {
+                    "type": "string",
+                    "description": (
+                        "All relevant background information, prior facts, decisions, and state needed by the agent to solve the current query. Should be as detailed and self-contained as possible."
+                    ),
+                    "required": True,
+                },
+            },
+        }
         super().__init__()
         self.function_name = "agent"
         self.tools = []
@@ -79,7 +70,6 @@ class AgentParam(LLMParam, ToolParamBase):
         self.custom_header = {}
 
 
-
 class Agent(LLM, ToolBase):
     component_name = "Agent"
 
@@ -91,13 +81,15 @@ class Agent(LLM, ToolBase):
             original_name = cpn.get_meta()["function"]["name"]
             indexed_name = f"{original_name}_{idx}"
             self.tools[indexed_name] = cpn
-
-        self.chat_mdl = LLMBundle(self._canvas.get_tenant_id(), TenantLLMService.llm_id2llm_type(self._param.llm_id), self._param.llm_id,
-                                  max_retries=self._param.max_retries,
-                                  retry_interval=self._param.delay_after_error,
-                                  max_rounds=self._param.max_rounds,
-                                  verbose_tool_use=True
-                                  )
+        chat_model_config = get_model_config_by_type_and_name(self._canvas.get_tenant_id(), TenantLLMService.llm_id2llm_type(self._param.llm_id), self._param.llm_id)
+        self.chat_mdl = LLMBundle(
+            self._canvas.get_tenant_id(),
+            chat_model_config,
+            max_retries=self._param.max_retries,
+            retry_interval=self._param.delay_after_error,
+            max_rounds=self._param.max_rounds,
+            verbose_tool_use=False,
+        )
         self.tool_meta = []
         for indexed_name, tool_obj in self.tools.items():
             original_meta = tool_obj.get_meta()
@@ -114,10 +106,30 @@ class Agent(LLM, ToolBase):
                 self.tools[tnm] = tool_call_session
         self.callback = partial(self._canvas.tool_use_callback, id)
         self.toolcall_session = LLMToolPluginCallSession(self.tools, self.callback)
-        #self.chat_mdl.bind_tools(self.toolcall_session, self.tool_metas)
+        if self.tool_meta:
+            self.chat_mdl.bind_tools(self.toolcall_session, self.tool_meta)
+
+    def _fit_messages(self, prompt: str, msg: list[dict]) -> list[dict]:
+        _, fitted_messages = message_fit_in(
+            [{"role": "system", "content": prompt}, *msg],
+            int(self.chat_mdl.max_length * 0.97),
+        )
+        return fitted_messages
+
+    @staticmethod
+    def _append_system_prompt(msg: list[dict], extra_prompt: str) -> None:
+        if extra_prompt and msg and msg[0]["role"] == "system":
+            msg[0]["content"] += "\n" + extra_prompt
+
+    @staticmethod
+    def _clean_formatted_answer(ans: str) -> str:
+        ans = re.sub(r"^.*</think>", "", ans, flags=re.DOTALL)
+        ans = re.sub(r"^.*```json", "", ans, flags=re.DOTALL)
+        return re.sub(r"```\n*$", "", ans, flags=re.DOTALL)
 
     def _load_tool_obj(self, cpn: dict) -> object:
         from agent.component import component_class
+
         tool_name = cpn["component_name"]
         param = component_class(tool_name + "Param")()
         param.update(cpn["params"])
@@ -130,7 +142,7 @@ class Agent(LLM, ToolBase):
         return component_class(cpn["component_name"])(self._canvas, cpn_id, param)
 
     def get_meta(self) -> dict[str, Any]:
-        self._param.function_name= self._id.split("-->")[-1]
+        self._param.function_name = self._id.split("-->")[-1]
         m = super().get_meta()
         if hasattr(self._param, "user_prompt") and self._param.user_prompt:
             m["function"]["parameters"]["properties"]["user_prompt"] = self._param.user_prompt
@@ -139,10 +151,7 @@ class Agent(LLM, ToolBase):
     def get_input_form(self) -> dict[str, dict]:
         res = {}
         for k, v in self.get_input_elements().items():
-            res[k] = {
-                "type": "line",
-                "name": v["name"]
-            }
+            res[k] = {"type": "line", "name": v["name"]}
         for cpn in self._param.tools:
             if not isinstance(cpn, LLM):
                 continue
@@ -175,7 +184,7 @@ class Agent(LLM, ToolBase):
     def _invoke(self, **kwargs):
         return asyncio.run(self._invoke_async(**kwargs))
 
-    @timeout(int(os.environ.get("COMPONENT_EXEC_TIMEOUT", 20*60)))
+    @timeout(int(os.environ.get("COMPONENT_EXEC_TIMEOUT", 20 * 60)))
     async def _invoke_async(self, **kwargs):
         if self.check_if_canceled("Agent processing"):
             return
@@ -204,19 +213,17 @@ class Agent(LLM, ToolBase):
             schema = json.dumps(output_schema, ensure_ascii=False, indent=2)
             schema_prompt = structured_output_prompt(schema)
 
-        downstreams = self._canvas.get_component(self._id)["downstream"] if self._canvas.get_component(self._id) else []
+        component = self._canvas.get_component(self._id)
+        downstreams = component["downstream"] if component else []
         ex = self.exception_handler()
-        if any([self._canvas.get_component_obj(cid).component_name.lower()=="message" for cid in downstreams]) and not (ex and ex["goto"]) and not output_schema:
+        has_message_downstream = any(self._canvas.get_component_obj(cid).component_name.lower() == "message" for cid in downstreams)
+        if has_message_downstream and not (ex and ex["goto"]) and not output_schema:
             self.set_output("content", partial(self.stream_output_with_tools_async, prompt, deepcopy(msg), user_defined_prompt))
             return
 
-        _, msg = message_fit_in([{"role": "system", "content": prompt}, *msg], int(self.chat_mdl.max_length * 0.97))
-        use_tools = []
-        ans = ""
-        async for delta_ans, _tk in self._react_with_tools_streamly_async_simple(prompt, msg, use_tools, user_defined_prompt,schema_prompt=schema_prompt):
-            if self.check_if_canceled("Agent processing"):
-                return
-            ans += delta_ans
+        msg = self._fit_messages(prompt, msg)
+        self._append_system_prompt(msg, schema_prompt)
+        ans = await self._generate_async(msg)
 
         if ans.find("**ERROR**") >= 0:
             logging.error(f"Agent._chat got error. response: {ans}")
@@ -230,14 +237,8 @@ class Agent(LLM, ToolBase):
             error = ""
             for _ in range(self._param.max_retries + 1):
                 try:
-                    def clean_formated_answer(ans: str) -> str:
-                        ans = re.sub(r"^.*</think>", "", ans, flags=re.DOTALL)
-                        ans = re.sub(r"^.*```json", "", ans, flags=re.DOTALL)
-                        return re.sub(r"```\n*$", "", ans, flags=re.DOTALL)
-                    obj = json_repair.loads(clean_formated_answer(ans))
+                    obj = json_repair.loads(self._clean_formatted_answer(ans))
                     self.set_output("structured", obj)
-                    if use_tools:
-                        self.set_output("use_tools", use_tools)
                     return obj
                 except Exception:
                     error = "The answer cannot be parsed as JSON"
@@ -248,330 +249,92 @@ class Agent(LLM, ToolBase):
             self.set_output("_ERROR", error)
             return
 
+        artifact_md = self._collect_tool_artifact_markdown(existing_text=ans)
+        if artifact_md:
+            ans += "\n\n" + artifact_md
         self.set_output("content", ans)
-        if use_tools:
-            self.set_output("use_tools", use_tools)
         return ans
 
     async def stream_output_with_tools_async(self, prompt, msg, user_defined_prompt={}):
-        _, msg = message_fit_in([{"role": "system", "content": prompt}, *msg], int(self.chat_mdl.max_length * 0.97))
-        answer_without_toolcall = ""
-        use_tools = []
-        async for delta_ans, _ in self._react_with_tools_streamly_async_simple(prompt, msg, use_tools, user_defined_prompt):
+        if len(msg) > 3:
+            st = timer()
+            user_request = await full_question(messages=msg, chat_mdl=self.chat_mdl)
+            self.callback("Multi-turn conversation optimization", {}, user_request, elapsed_time=timer() - st)
+            msg = [*msg[:-1], {"role": "user", "content": user_request}]
+
+        msg = self._fit_messages(prompt, msg)
+
+        need2cite = self._param.cite and self._canvas.get_reference()["chunks"] and self._id.find("-->") < 0
+        cited = False
+        if need2cite and len(msg) < 7:
+            self._append_system_prompt(msg, citation_prompt())
+            cited = True
+
+        answer = ""
+        async for delta in self._generate_streamly(msg):
             if self.check_if_canceled("Agent streaming"):
                 return
-
-            if delta_ans.find("**ERROR**") >= 0:
+            if delta.find("**ERROR**") >= 0:
                 if self.get_exception_default_value():
                     self.set_output("content", self.get_exception_default_value())
                     yield self.get_exception_default_value()
                 else:
-                    self.set_output("_ERROR", delta_ans)
-                    return
-            answer_without_toolcall += delta_ans
-            yield delta_ans
-
-        self.set_output("content", answer_without_toolcall)
-        if use_tools:
-            self.set_output("use_tools", use_tools)
-
-    async def _react_with_tools_streamly_async_simple(self, prompt, history: list[dict], use_tools, user_defined_prompt={}, schema_prompt: str = ""):
-        token_count = 0
-        tool_metas = self.tool_meta
-        hist = deepcopy(history)
-        last_calling = ""
-        if len(hist) > 3:
-            st = timer()
-            user_request = await full_question(messages=history, chat_mdl=self.chat_mdl)
-            self.callback("Multi-turn conversation optimization", {}, user_request, elapsed_time=timer()-st)
-        else:
-            user_request = history[-1]["content"]
-
-        def build_task_desc(prompt: str, user_request: str, user_defined_prompt: dict | None = None) -> str:
-            """Build a minimal task_desc by concatenating prompt, query, and tool schemas."""
-            user_defined_prompt = user_defined_prompt or {}
-
-            task_desc = (
-                "### Agent Prompt\n"
-                f"{prompt}\n\n"
-                "### User Request\n"
-                f"{user_request}\n\n"
-            )
-
-            if user_defined_prompt:
-                udp_json = json.dumps(user_defined_prompt, ensure_ascii=False, indent=2)
-                task_desc += "\n### User Defined Prompts\n" + udp_json + "\n"
-
-            return task_desc
-
-
-        async def use_tool_async(name, args):
-            nonlocal hist, use_tools, last_calling
-            logging.info(f"{last_calling=} == {name=}")
-            last_calling = name
-            tool_response = await self.toolcall_session.tool_call_async(name, args)
-            use_tools.append({
-                "name": name,
-                "arguments": args,
-                "results": tool_response
-            })
-            return name, tool_response
-
-        async def complete():
-            nonlocal hist
-            need2cite = self._param.cite and self._canvas.get_reference()["chunks"] and self._id.find("-->") < 0
-            if schema_prompt:
-                need2cite = False
-            cited = False
-            if hist and hist[0]["role"] == "system":
-                if schema_prompt:
-                    hist[0]["content"] += "\n" + schema_prompt
-                if need2cite and len(hist) < 7:
-                    hist[0]["content"] += citation_prompt()
-                    cited = True
-            yield "", token_count
-
-            _hist = hist
-            if len(hist) > 12:
-                _hist = [hist[0], hist[1], *hist[-10:]]
-            entire_txt = ""
-            async for delta_ans in self._generate_streamly(_hist):
-                if not need2cite or cited:
-                    yield delta_ans, 0
-                entire_txt += delta_ans
-            if not need2cite or cited:
+                    self.set_output("_ERROR", delta)
                 return
+            if not need2cite or cited:
+                yield delta
+            answer += delta
 
-            st = timer()
-            txt = ""
-            async for delta_ans in self._gen_citations_async(entire_txt):
-                if self.check_if_canceled("Agent streaming"):
-                    return
-                yield delta_ans, 0
-                txt += delta_ans
-
-            self.callback("gen_citations", {}, txt, elapsed_time=timer()-st)
-
-        def build_observation(tool_call_res: list[tuple]) -> str:
-            """
-            Build a Observation from tool call results.
-            No LLM involved.
-            """
-            if not tool_call_res:
-                return ""
-
-            lines = ["Observation:"]
-            for name, result in tool_call_res:
-                lines.append(f"[{name} result]")
-                lines.append(str(result))
-
-            return "\n".join(lines)
-
-        def append_user_content(hist, content):
-            if hist[-1]["role"] == "user":
-                hist[-1]["content"] += content
-            else:
-                hist.append({"role": "user", "content": content})
+        if not need2cite or cited:
+            artifact_md = self._collect_tool_artifact_markdown(existing_text=answer)
+            if artifact_md:
+                yield "\n\n" + artifact_md
+                answer += "\n\n" + artifact_md
+            self.set_output("content", answer)
+            return
 
         st = timer()
-        task_desc = build_task_desc(prompt, user_request, user_defined_prompt)
-        self.callback("analyze_task", {}, task_desc, elapsed_time=timer()-st)
-        for _ in range(self._param.max_rounds + 1):
+        cited_answer = ""
+        async for delta in self._gen_citations_async(answer):
             if self.check_if_canceled("Agent streaming"):
                 return
-            response, tk = await next_step_async(self.chat_mdl, hist, tool_metas, task_desc, user_defined_prompt)
-            # self.callback("next_step", {}, str(response)[:256]+"...")
-            token_count += tk or 0
-            hist.append({"role": "assistant", "content": response})
-            try:
-                functions = json_repair.loads(re.sub(r"```.*", "", response))
-                if not isinstance(functions, list):
-                    raise TypeError(f"List should be returned, but `{functions}`")
-                for f in functions:
-                    if not isinstance(f, dict):
-                        raise TypeError(f"An object type should be returned, but `{f}`")
-
-                tool_tasks = []
-                for func in functions:
-                    name = func["name"]
-                    args = func["arguments"]
-                    if name == COMPLETE_TASK:
-                        append_user_content(hist, f"Respond with a formal answer. FORGET(DO NOT mention) about `{COMPLETE_TASK}`. The language for the response MUST be as the same as the first user request.\n")
-                        async for txt, tkcnt in complete():
-                            yield txt, tkcnt
-                        return
-
-                    tool_tasks.append(asyncio.create_task(use_tool_async(name, args)))
-
-                results = await asyncio.gather(*tool_tasks) if tool_tasks else []
-                st = timer()
-                reflection = build_observation(results)
-                append_user_content(hist, reflection)
-                self.callback("reflection", {}, str(reflection), elapsed_time=timer()-st)
-
-            except Exception as e:
-                logging.exception(msg=f"Wrong JSON argument format in LLM ReAct response: {e}")
-                e = f"\nTool call error, please correct the input parameter of response format and call it again.\n *** Exception ***\n{e}"
-                append_user_content(hist, str(e))
-
-        logging.warning( f"Exceed max rounds: {self._param.max_rounds}")
-        final_instruction = f"""
-{user_request}
-IMPORTANT: You have reached the conversation limit. Based on ALL the information and research you have gathered so far, please provide a DIRECT and COMPREHENSIVE final answer to the original request.
-Instructions:
-1. SYNTHESIZE all information collected during this conversation
-2. Provide a COMPLETE response using existing data - do not suggest additional research
-3. Structure your response as a FINAL DELIVERABLE, not a plan
-4. If information is incomplete, state what you found and provide the best analysis possible with available data
-5. DO NOT mention conversation limits or suggest further steps
-6. Focus on delivering VALUE with the information already gathered
-Respond immediately with your final comprehensive answer.
-        """
-        if self.check_if_canceled("Agent final instruction"):
-            return
-        append_user_content(hist, final_instruction)
-
-        async for txt, tkcnt in complete():
-            yield txt, tkcnt
-
-#     async def _react_with_tools_streamly_async(self, prompt, history: list[dict], use_tools, user_defined_prompt={}, schema_prompt: str = ""):
-#         token_count = 0
-#         tool_metas = self.tool_meta
-#         hist = deepcopy(history)
-#         last_calling = ""
-#         if len(hist) > 3:
-#             st = timer()
-#             user_request = await full_question(messages=history, chat_mdl=self.chat_mdl)
-#             self.callback("Multi-turn conversation optimization", {}, user_request, elapsed_time=timer()-st)
-#         else:
-#             user_request = history[-1]["content"]
-
-#         async def use_tool_async(name, args):
-#             nonlocal hist, use_tools, last_calling
-#             logging.info(f"{last_calling=} == {name=}")
-#             last_calling = name
-#             tool_response = await self.toolcall_session.tool_call_async(name, args)
-#             use_tools.append({
-#                 "name": name,
-#                 "arguments": args,
-#                 "results": tool_response
-#             })
-#             # self.callback("add_memory", {}, "...")
-#             #self.add_memory(hist[-2]["content"], hist[-1]["content"], name, args, str(tool_response), user_defined_prompt)
-
-#             return name, tool_response
-
-#         async def complete():
-#             nonlocal hist
-#             need2cite = self._param.cite and self._canvas.get_reference()["chunks"] and self._id.find("-->") < 0
-#             if schema_prompt:
-#                 need2cite = False
-#             cited = False
-#             if hist and hist[0]["role"] == "system":
-#                 if schema_prompt:
-#                     hist[0]["content"] += "\n" + schema_prompt
-#                 if need2cite and len(hist) < 7:
-#                     hist[0]["content"] += citation_prompt()
-#                     cited = True
-#             yield "", token_count
-
-#             _hist = hist
-#             if len(hist) > 12:
-#                 _hist = [hist[0], hist[1], *hist[-10:]]
-#             entire_txt = ""
-#             async for delta_ans in self._generate_streamly(_hist):
-#                 if not need2cite or cited:
-#                     yield delta_ans, 0
-#                 entire_txt += delta_ans
-#             if not need2cite or cited:
-#                 return
-
-#             st = timer()
-#             txt = ""
-#             async for delta_ans in self._gen_citations_async(entire_txt):
-#                 if self.check_if_canceled("Agent streaming"):
-#                     return
-#                 yield delta_ans, 0
-#                 txt += delta_ans
-
-#             self.callback("gen_citations", {}, txt, elapsed_time=timer()-st)
-
-#         def append_user_content(hist, content):
-#             if hist[-1]["role"] == "user":
-#                 hist[-1]["content"] += content
-#             else:
-#                 hist.append({"role": "user", "content": content})
-
-#         st = timer()
-#         task_desc = await analyze_task_async(self.chat_mdl, prompt, user_request, tool_metas, user_defined_prompt)
-#         self.callback("analyze_task", {}, task_desc, elapsed_time=timer()-st)
-#         for _ in range(self._param.max_rounds + 1):
-#             if self.check_if_canceled("Agent streaming"):
-#                 return
-#             response, tk = await next_step_async(self.chat_mdl, hist, tool_metas, task_desc, user_defined_prompt)
-#             # self.callback("next_step", {}, str(response)[:256]+"...")
-#             token_count += tk or 0
-#             hist.append({"role": "assistant", "content": response})
-#             try:
-#                 functions = json_repair.loads(re.sub(r"```.*", "", response))
-#                 if not isinstance(functions, list):
-#                     raise TypeError(f"List should be returned, but `{functions}`")
-#                 for f in functions:
-#                     if not isinstance(f, dict):
-#                         raise TypeError(f"An object type should be returned, but `{f}`")
-
-#                 tool_tasks = []
-#                 for func in functions:
-#                     name = func["name"]
-#                     args = func["arguments"]
-#                     if name == COMPLETE_TASK:
-#                         append_user_content(hist, f"Respond with a formal answer. FORGET(DO NOT mention) about `{COMPLETE_TASK}`. The language for the response MUST be as the same as the first user request.\n")
-#                         async for txt, tkcnt in complete():
-#                             yield txt, tkcnt
-#                         return
-
-#                     tool_tasks.append(asyncio.create_task(use_tool_async(name, args)))
-
-#                 results = await asyncio.gather(*tool_tasks) if tool_tasks else []
-#                 st = timer()
-#                 reflection = await reflect_async(self.chat_mdl, hist, results, user_defined_prompt)
-#                 append_user_content(hist, reflection)
-#                 self.callback("reflection", {}, str(reflection), elapsed_time=timer()-st)
-
-#             except Exception as e:
-#                 logging.exception(msg=f"Wrong JSON argument format in LLM ReAct response: {e}")
-#                 e = f"\nTool call error, please correct the input parameter of response format and call it again.\n *** Exception ***\n{e}"
-#                 append_user_content(hist, str(e))
-
-#         logging.warning( f"Exceed max rounds: {self._param.max_rounds}")
-#         final_instruction = f"""
-# {user_request}
-# IMPORTANT: You have reached the conversation limit. Based on ALL the information and research you have gathered so far, please provide a DIRECT and COMPREHENSIVE final answer to the original request.
-# Instructions:
-# 1. SYNTHESIZE all information collected during this conversation
-# 2. Provide a COMPLETE response using existing data - do not suggest additional research
-# 3. Structure your response as a FINAL DELIVERABLE, not a plan
-# 4. If information is incomplete, state what you found and provide the best analysis possible with available data
-# 5. DO NOT mention conversation limits or suggest further steps
-# 6. Focus on delivering VALUE with the information already gathered
-# Respond immediately with your final comprehensive answer.
-#         """
-#         if self.check_if_canceled("Agent final instruction"):
-#             return
-#         append_user_content(hist, final_instruction)
-
-#         async for txt, tkcnt in complete():
-#             yield txt, tkcnt
+            yield delta
+            cited_answer += delta
+        artifact_md = self._collect_tool_artifact_markdown(existing_text=cited_answer)
+        if artifact_md:
+            yield "\n\n" + artifact_md
+            cited_answer += "\n\n" + artifact_md
+        self.callback("gen_citations", {}, cited_answer, elapsed_time=timer() - st)
+        self.set_output("content", cited_answer)
 
     async def _gen_citations_async(self, text):
         retrievals = self._canvas.get_reference()
         retrievals = {"chunks": list(retrievals["chunks"].values()), "doc_aggs": list(retrievals["doc_aggs"].values())}
         formated_refer = kb_prompt(retrievals, self.chat_mdl.max_length, True)
-        async for delta_ans in self._generate_streamly([{"role": "system", "content": citation_plus("\n\n".join(formated_refer))},
-                                                  {"role": "user", "content": text}
-                                                  ]):
+        async for delta_ans in self._generate_streamly([{"role": "system", "content": citation_plus("\n\n".join(formated_refer))}, {"role": "user", "content": text}]):
             yield delta_ans
 
+    def _collect_tool_artifact_markdown(self, existing_text: str = "") -> str:
+        md_parts = []
+        for tool_obj in self.tools.values():
+            if not hasattr(tool_obj, "_param") or not hasattr(tool_obj._param, "outputs"):
+                continue
+            artifacts_meta = tool_obj._param.outputs.get("_ARTIFACTS", {})
+            artifacts = artifacts_meta.get("value") if isinstance(artifacts_meta, dict) else None
+            if not artifacts:
+                continue
+            for art in artifacts:
+                if not isinstance(art, dict):
+                    continue
+                url = art.get("url", "")
+                if url and (f"![]({url})" in existing_text or f"![{art.get('name', '')}]({url})" in existing_text):
+                    continue
+                if art.get("mime_type", "").startswith("image/"):
+                    md_parts.append(f"![{art['name']}]({url})")
+                else:
+                    md_parts.append(f"[Download {art['name']}]({url})")
+        return "\n\n".join(md_parts)
+
     def reset(self, only_output=False):
         """
         Reset all tools if they have a reset method. This avoids errors for tools like MCPToolCallSession.

agent/component/begin.py

@@ -41,6 +41,7 @@ class Begin(UserFillUp):
         if self.check_if_canceled("Begin processing"):
             return
 
+        layout_recognize = self._param.layout_recognize or None
         for k, v in kwargs.get("inputs", {}).items():
             if self.check_if_canceled("Begin processing"):
                 return
@@ -52,7 +53,7 @@ class Begin(UserFillUp):
                     file_value = v["value"]
                     # Support both single file (backward compatibility) and multiple files
                     files = file_value if isinstance(file_value, list) else [file_value]
-                    v = FileService.get_files(files)
+                    v = FileService.get_files(files, layout_recognize=layout_recognize)
             else:
                 v = v.get("value")
             self.set_output(k, v)

agent/component/categorize.py

@@ -21,6 +21,7 @@ from abc import ABC
 
 from common.constants import LLMType
 from api.db.services.llm_service import LLMBundle
+from api.db.joint_services.tenant_model_service import get_model_config_by_type_and_name
 from agent.component.llm import LLMParam, LLM
 from common.connection_utils import timeout
 from rag.llm.chat_model import ERROR_PREFIX
@@ -122,7 +123,8 @@ class Categorize(LLM, ABC):
         msg[-1]["content"] = query_value
         self.set_input_value(query_key, msg[-1]["content"])
         self._param.update_prompt()
-        chat_mdl = LLMBundle(self._canvas.get_tenant_id(), LLMType.CHAT, self._param.llm_id)
+        chat_model_config = get_model_config_by_type_and_name(self._canvas.get_tenant_id(), LLMType.CHAT, self._param.llm_id)
+        chat_mdl = LLMBundle(self._canvas.get_tenant_id(), chat_model_config)
 
         user_prompt = """
 ---- Real Data ----

agent/component/data_operations.py

@@ -94,9 +94,9 @@ class DataOperations(ComponentBase,ABC):
 
     def _recursive_eval(self, data):
         if isinstance(data, dict):
-            return {k: self.recursive_eval(v) for k, v in data.items()}
+            return {k: self._recursive_eval(v) for k, v in data.items()}
         if isinstance(data, list):
-            return [self.recursive_eval(item) for item in data]
+            return [self._recursive_eval(item) for item in data]
         if isinstance(data, str):
             try:
                 if (

agent/component/fillup.py

@@ -27,6 +27,7 @@ class UserFillUpParam(ComponentParamBase):
         super().__init__()
         self.enable_tips = True
         self.tips = "Please fill up the form"
+        self.layout_recognize = ""
 
     def check(self) -> bool:
         return True
@@ -61,6 +62,7 @@ class UserFillUp(ComponentBase):
                 content = re.sub(r"\{%s\}"%k, ans, content)
 
             self.set_output("tips", content)
+        layout_recognize = self._param.layout_recognize or None
         for k, v in kwargs.get("inputs", {}).items():
             if self.check_if_canceled("UserFillUp processing"):
                 return
@@ -71,7 +73,7 @@ class UserFillUp(ComponentBase):
                     file_value = v["value"]
                     # Support both single file (backward compatibility) and multiple files
                     files = file_value if isinstance(file_value, list) else [file_value]
-                    v = FileService.get_files(files)
+                    v = FileService.get_files(files, layout_recognize=layout_recognize)
             else:
                 v = v.get("value")
             self.set_output(k, v)

agent/component/invoke.py

@@ -19,6 +19,7 @@ import os
 import re
 import time
 from abc import ABC
+from functools import partial
 
 import requests
 
@@ -29,7 +30,7 @@ from deepdoc.parser import HtmlParser
 
 class InvokeParam(ComponentParamBase):
     """
-    Define the Crawler component parameters.
+    Define the Invoke component parameters.
     """
 
     def __init__(self):
@@ -41,7 +42,7 @@ class InvokeParam(ComponentParamBase):
         self.url = ""
         self.timeout = 60
         self.clean_html = False
-        self.datatype = "json"  # New parameter to determine data posting type
+        self.datatype = "json"
 
     def check(self):
         self.check_valid_value(self.method.lower(), "Type of content from the crawler", ["get", "post", "put"])
@@ -53,92 +54,199 @@ class InvokeParam(ComponentParamBase):
 
 class Invoke(ComponentBase, ABC):
     component_name = "Invoke"
+    header_variable_ref_patt = r"\{([a-zA-Z_][a-zA-Z0-9_.@-]*)\}"
 
+    @staticmethod
+    def _coerce_json_arg_if_possible(key, value):
+        raw_value = value
+        if isinstance(value, str):
+            try:
+                value = json.loads(value)
+                logging.debug(
+                    "Invoke JSON arg coercion succeeded. key=%s parsed_type=%s",
+                    key,
+                    type(value).__name__,
+                )
+            except json.JSONDecodeError as exc:
+                logging.info(
+                    "Invoke JSON arg coercion skipped; value is not valid JSON. key=%s raw=%r error=%s",
+                    key,
+                    raw_value,
+                    exc,
+                )
+                return raw_value
+
+        try:
+            json.dumps(value, allow_nan=False)
+        except (TypeError, ValueError) as exc:
+            logging.warning(
+                "Invoke JSON arg is not JSON-serializable. key=%s value_type=%s value=%r error=%s",
+                key,
+                type(value).__name__,
+                value,
+                exc,
+            )
+            raise ValueError(f"Invoke JSON argument '{key}' is not JSON-serializable.") from exc
+
+        return value
+
+    def get_input_form(self) -> dict[str, dict]:
+        res = {}
+        for item in self._param.variables or []:
+            if not isinstance(item, dict):
+                continue
+            ref = (item.get("ref") or "").strip()
+            if not ref or ref in res:
+                continue
+
+            elements = self.get_input_elements_from_text("{" + ref + "}")
+            element = elements.get(ref, {})
+            res[ref] = {
+                "type": "line",
+                "name": element.get("name") or item.get("key") or ref,
+            }
+        return res
+
+    def _resolve_variable_value(self, variable_name: str, kwargs: dict | None = None):
+        kwargs = kwargs or {}
+        value = kwargs.get(variable_name, self._canvas.get_variable_value(variable_name))
+        if isinstance(value, partial):
+            value = "".join(value())
+            self.set_input_value(variable_name, value)
+        return "" if value is None else value
+
+    def _render_template(self, content: str, pattern: str, kwargs: dict | None = None, *, flags: int = 0) -> str:
+        content = content or ""
+        if not content:
+            return content
+
+        def replace_variable(match_obj):
+            return str(self._resolve_variable_value(match_obj.group(1), kwargs))
+
+        return re.sub(pattern, replace_variable, content, flags=flags)
+
+    def _resolve_template_text(self, content: str, kwargs: dict | None = None) -> str:
+        return self._render_template(content, self.variable_ref_patt, kwargs, flags=re.DOTALL)
+
+    def _resolve_header_text(self, content: str, kwargs: dict | None = None) -> str:
+        # Headers support plain {token} placeholders, so they cannot reuse the canvas variable regex.
+        return self._render_template(content, self.header_variable_ref_patt, kwargs)
+
+    def _resolve_arg_value(self, para: dict, kwargs: dict) -> object:
+        ref = (para.get("ref") or "").strip()
+        if ref and (ref in kwargs or self._canvas.get_variable_value(ref) is not None):
+            return self._resolve_variable_value(ref, kwargs)
+
+        if para.get("value") is not None:
+            value = para["value"]
+            if isinstance(value, str):
+                return self._resolve_template_text(value, kwargs)
+            return value
+
+        if ref:
+            return self._resolve_variable_value(ref, kwargs)
+
+        return ""
+
+    def _is_json_mode(self) -> bool:
+        return self._param.datatype.lower() == "json"
+
+    def _build_request_args(self, kwargs: dict) -> dict:
+        args = {}
+        for para in self._param.variables:
+            key = para["key"]
+            value = self._resolve_arg_value(para, kwargs)
+            if self._is_json_mode():
+                # JSON mode accepts stringified JSON so complex payloads can be passed through variables.
+                value = self._coerce_json_arg_if_possible(key, value)
+            args[key] = value
+
+            if para.get("ref"):
+                self.set_input_value(para["ref"], value)
+        return args
+
+    def _build_url(self, kwargs: dict) -> str:
+        url = self._resolve_template_text(self._param.url.strip(), kwargs)
+        if not url.startswith(("http://", "https://")):
+            url = "http://" + url
+        return url
+
+    def _build_headers(self, kwargs: dict) -> dict:
+        if not self._param.headers:
+            return {}
+
+        headers = json.loads(self._param.headers)
+        if not isinstance(headers, dict):
+            raise ValueError("Invoke headers must be a JSON object.")
+
+        return {
+            key: self._resolve_header_text(value, kwargs) if isinstance(value, str) else value
+            for key, value in headers.items()
+        }
+
+    def _build_proxies(self) -> dict | None:
+        if not re.sub(r"https?:?/?/?", "", self._param.proxy):
+            return None
+        return {"http": self._param.proxy, "https": self._param.proxy}
+
+    def _send_request(self, url: str, args: dict, headers: dict, proxies: dict | None):
+        method = self._param.method.lower()
+        request = getattr(requests, method)
+        request_kwargs = {
+            "url": url,
+            "headers": headers,
+            "proxies": proxies,
+            "timeout": self._param.timeout,
+        }
+
+        # GET sends query params; POST/PUT send either JSON or form data based on datatype.
+        if method == "get":
+            request_kwargs["params"] = args
+            return request(**request_kwargs)
+
+        body_key = "json" if self._is_json_mode() else "data"
+        request_kwargs[body_key] = args
+        return request(**request_kwargs)
+
+    def _format_response(self, response) -> str:
+        if not self._param.clean_html:
+            return response.text
+
+        # HtmlParser keeps the Invoke output text-focused when the endpoint returns HTML.
+        sections = HtmlParser()(None, response.content)
+        return "\n".join(sections)
+    
     @timeout(int(os.environ.get("COMPONENT_EXEC_TIMEOUT", 3)))
     def _invoke(self, **kwargs):
         if self.check_if_canceled("Invoke processing"):
             return
 
-        args = {}
-        for para in self._param.variables:
-            if para.get("value"):
-                args[para["key"]] = para["value"]
-            else:
-                args[para["key"]] = self._canvas.get_variable_value(para["ref"])
+        args = self._build_request_args(kwargs)
+        url = self._build_url(kwargs)
+        headers = self._build_headers(kwargs)
+        proxies = self._build_proxies()
 
-        url = self._param.url.strip()
-
-        def replace_variable(match):
-            var_name = match.group(1)
-            try:
-                value = self._canvas.get_variable_value(var_name)
-                return str(value or "")
-            except Exception:
-                return ""
-
-        # {base_url} or {component_id@variable_name}
-        url = re.sub(r"\{([a-zA-Z_][a-zA-Z0-9_.@-]*)\}", replace_variable, url)
-
-        if url.find("http") != 0:
-            url = "http://" + url
-
-        method = self._param.method.lower()
-        headers = {}
-        if self._param.headers:
-            headers = json.loads(self._param.headers)
-        proxies = None
-        if re.sub(r"https?:?/?/?", "", self._param.proxy):
-            proxies = {"http": self._param.proxy, "https": self._param.proxy}
-
-        last_e = ""
+        last_error = None
         for _ in range(self._param.max_retries + 1):
             if self.check_if_canceled("Invoke processing"):
                 return
 
             try:
-                if method == "get":
-                    response = requests.get(url=url, params=args, headers=headers, proxies=proxies, timeout=self._param.timeout)
-                    if self._param.clean_html:
-                        sections = HtmlParser()(None, response.content)
-                        self.set_output("result", "\n".join(sections))
-                    else:
-                        self.set_output("result", response.text)
-
-                if method == "put":
-                    if self._param.datatype.lower() == "json":
-                        response = requests.put(url=url, json=args, headers=headers, proxies=proxies, timeout=self._param.timeout)
-                    else:
-                        response = requests.put(url=url, data=args, headers=headers, proxies=proxies, timeout=self._param.timeout)
-                    if self._param.clean_html:
-                        sections = HtmlParser()(None, response.content)
-                        self.set_output("result", "\n".join(sections))
-                    else:
-                        self.set_output("result", response.text)
-
-                if method == "post":
-                    if self._param.datatype.lower() == "json":
-                        response = requests.post(url=url, json=args, headers=headers, proxies=proxies, timeout=self._param.timeout)
-                    else:
-                        response = requests.post(url=url, data=args, headers=headers, proxies=proxies, timeout=self._param.timeout)
-                    if self._param.clean_html:
-                        self.set_output("result", "\n".join(sections))
-                    else:
-                        self.set_output("result", response.text)
-
-                return self.output("result")
+                response = self._send_request(url, args, headers, proxies)
+                result = self._format_response(response)
+                self.set_output("result", result)
+                return result
             except Exception as e:
                 if self.check_if_canceled("Invoke processing"):
                     return
 
-                last_e = e
+                last_error = e
                 logging.exception(f"Http request error: {e}")
                 time.sleep(self._param.delay_after_error)
 
-        if last_e:
-            self.set_output("_ERROR", str(last_e))
-            return f"Http request error: {last_e}"
-
-        assert False, self.output()
+        if last_error:
+            self.set_output("_ERROR", str(last_error))
+            return f"Http request error: {last_error}"
 
     def thoughts(self) -> str:
         return "Waiting for the server respond..."

agent/component/iterationitem.py

@@ -69,7 +69,7 @@ class IterationItem(ComponentBase, ABC):
             if p._id != pid:
                 continue
 
-            if p.component_name.lower() in ["categorize", "message", "switch", "userfillup", "interationitem"]:
+            if p.component_name.lower() in ["categorize", "message", "switch", "userfillup", "iterationitem"]:
                 continue
 
             for k, o in p._param.outputs.items():

agent/component/llm.py

@@ -25,6 +25,7 @@ from functools import partial
 from common.constants import LLMType
 from api.db.services.llm_service import LLMBundle
 from api.db.services.tenant_llm_service import TenantLLMService
+from api.db.joint_services.tenant_model_service import get_model_config_by_type_and_name
 from agent.component.base import ComponentBase, ComponentParamBase
 from common.connection_utils import timeout
 from rag.prompts.generator import tool_call_summary, message_fit_in, citation_prompt, structured_output_prompt
@@ -84,10 +85,10 @@ class LLM(ComponentBase):
 
     def __init__(self, canvas, component_id, param: ComponentParamBase):
         super().__init__(canvas, component_id, param)
-        self.chat_mdl = LLMBundle(self._canvas.get_tenant_id(), TenantLLMService.llm_id2llm_type(self._param.llm_id),
-                                  self._param.llm_id, max_retries=self._param.max_retries,
-                                  retry_interval=self._param.delay_after_error
-                                  )
+        chat_model_config = get_model_config_by_type_and_name(self._canvas.get_tenant_id(), TenantLLMService.llm_id2llm_type(self._param.llm_id), self._param.llm_id)
+        self.chat_mdl = LLMBundle(self._canvas.get_tenant_id(), chat_model_config,
+                                  max_retries=self._param.max_retries,
+                                  retry_interval=self._param.delay_after_error)
         self.imgs = []
 
     def get_input_form(self) -> dict[str, dict]:
@@ -125,23 +126,119 @@ class LLM(ComponentBase):
             msg.append(p)
         return msg, self.string_format(self._param.sys_prompt, args)
 
-    def _prepare_prompt_variables(self):
-        if self._param.visual_files_var:
-            self.imgs = self._canvas.get_variable_value(self._param.visual_files_var)
-            if not self.imgs:
-                self.imgs = []
-            self.imgs = [img for img in self.imgs if img[:len("data:image/")] == "data:image/"]
-            if self.imgs and TenantLLMService.llm_id2llm_type(self._param.llm_id) == LLMType.CHAT.value:
-                self.chat_mdl = LLMBundle(self._canvas.get_tenant_id(), LLMType.IMAGE2TEXT.value,
-                                          self._param.llm_id, max_retries=self._param.max_retries,
-                                          retry_interval=self._param.delay_after_error
-                                          )
+    @staticmethod
+    def _extract_data_images(value) -> list[str]:
+        imgs = []
 
+        def walk(v):
+            if v is None:
+                return
+            if isinstance(v, str):
+                v = v.strip()
+                if v.startswith("data:image/"):
+                    imgs.append(v)
+                return
+            if isinstance(v, (list, tuple, set)):
+                for item in v:
+                    walk(item)
+                return
+            if isinstance(v, dict):
+                if "content" in v:
+                    walk(v.get("content"))
+                else:
+                    for item in v.values():
+                        walk(item)
+
+        walk(value)
+        return imgs
+
+    @staticmethod
+    def _uniq_images(images: list[str]) -> list[str]:
+        seen = set()
+        uniq = []
+        for img in images:
+            if not isinstance(img, str):
+                continue
+            if not img.startswith("data:image/"):
+                continue
+            if img in seen:
+                continue
+            seen.add(img)
+            uniq.append(img)
+        return uniq
+
+    @classmethod
+    def _remove_data_images(cls, value):
+        if value is None:
+            return None
+
+        if isinstance(value, str):
+            return None if value.strip().startswith("data:image/") else value
+
+        if isinstance(value, list):
+            cleaned = []
+            for item in value:
+                v = cls._remove_data_images(item)
+                if v is None:
+                    continue
+                if isinstance(v, (list, tuple, set, dict)) and not v:
+                    continue
+                cleaned.append(v)
+            return cleaned
+
+        if isinstance(value, tuple):
+            cleaned = []
+            for item in value:
+                v = cls._remove_data_images(item)
+                if v is None:
+                    continue
+                if isinstance(v, (list, tuple, set, dict)) and not v:
+                    continue
+                cleaned.append(v)
+            return tuple(cleaned)
+
+        if isinstance(value, set):
+            cleaned = []
+            for item in value:
+                v = cls._remove_data_images(item)
+                if v is None:
+                    continue
+                if isinstance(v, (list, tuple, set, dict)) and not v:
+                    continue
+                cleaned.append(v)
+            return cleaned
+
+        if isinstance(value, dict):
+            if value.get("type") in {"image_url", "input_image", "image"} and cls._extract_data_images(value):
+                return None
+
+            cleaned = {}
+            for k, item in value.items():
+                v = cls._remove_data_images(item)
+                if v is None:
+                    continue
+                if isinstance(v, (list, tuple, set, dict)) and not v:
+                    continue
+                cleaned[k] = v
+            return cleaned
+
+        return value
+
+    def _prepare_prompt_variables(self):
+        self.imgs = []
+        if self._param.visual_files_var:
+            visual_val = self._canvas.get_variable_value(self._param.visual_files_var)
+            self.imgs.extend(self._extract_data_images(visual_val))
 
         args = {}
         vars = self.get_input_elements() if not self._param.debug_inputs else self._param.debug_inputs
+        extracted_imgs = []
         for k, o in vars.items():
-            args[k] = o["value"]
+            raw_value = o["value"]
+            extracted_imgs.extend(self._extract_data_images(raw_value))
+            args[k] = self._remove_data_images(raw_value)
+            if args[k] is None:
+                args[k] = ""
             if not isinstance(args[k], str):
                 try:
                     args[k] = json.dumps(args[k], ensure_ascii=False)
@@ -149,6 +246,13 @@ class LLM(ComponentBase):
                     args[k] = str(args[k])
             self.set_input_value(k, args[k])
 
+        self.imgs = self._uniq_images(self.imgs + extracted_imgs)
+        if self.imgs and TenantLLMService.llm_id2llm_type(self._param.llm_id) == LLMType.CHAT.value:
+            self.chat_mdl = LLMBundle(self._canvas.get_tenant_id(), LLMType.IMAGE2TEXT.value,
+                                      self._param.llm_id, max_retries=self._param.max_retries,
+                                      retry_interval=self._param.delay_after_error
+                                      )
+
         msg, sys_prompt = self._sys_prompt_and_msg(self._canvas.get_history(self._param.message_history_window_size)[:-1], args)
         user_defined_prompt, sys_prompt = self._extract_prompts(sys_prompt)
         if self._param.cite and self._canvas.get_reference()["chunks"]:

agent/component/message.py

@@ -14,8 +14,11 @@
 #  limitations under the License.
 #
 import asyncio
-import nest_asyncio
-nest_asyncio.apply()
+try:
+    import nest_asyncio
+    nest_asyncio.apply()
+except Exception:
+    pass
 import inspect
 import json
 import os
@@ -27,7 +30,9 @@ from functools import partial
 from typing import Any
 
 from agent.component.base import ComponentBase, ComponentParamBase
-from jinja2 import Template as Jinja2Template
+from jinja2.sandbox import SandboxedEnvironment
+
+_jinja2_sandbox = SandboxedEnvironment()
 
 from common.connection_utils import timeout
 from common.misc_utils import get_uuid
@@ -49,6 +54,9 @@ class MessageParam(ComponentParamBase):
         self.outputs = {
             "content": {
                 "type": "str"
+            },
+            "downloads": {
+                "type": "list"
             }
         }
 
@@ -61,10 +69,66 @@ class MessageParam(ComponentParamBase):
 class Message(ComponentBase):
     component_name = "Message"
 
+    @staticmethod
+    def _is_download_info(value: Any) -> bool:
+        return isinstance(value, dict) and all(
+            key in value for key in ("doc_id", "filename", "mime_type")
+        )
+
+    def _extract_downloads(self, value: Any) -> list[dict[str, Any]]:
+        if isinstance(value, str):
+            try:
+                value = json.loads(value)
+            except Exception:
+                return []
+
+        if self._is_download_info(value):
+            return [value]
+
+        if isinstance(value, list) and all(self._is_download_info(item) for item in value):
+            return value
+
+        return []
+
+    def _stringify_message_value(
+        self,
+        value: Any,
+        delimiter: str = None,
+        downloads: list[dict[str, Any]] | None = None,
+        fallback_to_str: bool = False,
+    ) -> str:
+        extracted_downloads = self._extract_downloads(value)
+        if extracted_downloads:
+            if downloads is not None:
+                downloads.extend(extracted_downloads)
+            return ""
+
+        if value is None:
+            return ""
+
+        if isinstance(value, list) and delimiter:
+            return delimiter.join([str(vv) for vv in value])
+
+        if isinstance(value, str):
+            return value
+
+        try:
+            return json.dumps(value, ensure_ascii=False)
+        except Exception:
+            if fallback_to_str:
+                return str(value)
+            return ""
+
     def get_input_elements(self) -> dict[str, Any]:
         return self.get_input_elements_from_text("".join(self._param.content))
 
-    def get_kwargs(self, script:str, kwargs:dict = {}, delimiter:str=None) -> tuple[str, dict[str, str | list | Any]]:
+    def get_kwargs(
+        self,
+        script: str,
+        kwargs: dict = {},
+        delimiter: str = None,
+        downloads: list[dict[str, Any]] | None = None,
+    ) -> tuple[str, dict[str, str | list | Any]]:
         for k,v in self.get_input_elements_from_text(script).items():
             if k in kwargs:
                 continue
@@ -79,15 +143,8 @@ class Message(ComponentBase):
                 else:
                     for t in iter_obj:
                         ans += t
-            elif isinstance(v, list) and delimiter:
-                ans = delimiter.join([str(vv) for vv in v])
-            elif not isinstance(v, str):
-                try:
-                    ans = json.dumps(v, ensure_ascii=False)
-                except Exception:
-                    pass
             else:
-                ans = v
+                ans = self._stringify_message_value(v, delimiter, downloads)
             if not ans:
                 ans = ""
             kwargs[k] = ans
@@ -110,6 +167,7 @@ class Message(ComponentBase):
         s = 0
         all_content = ""
         cache = {}
+        downloads = []
         for r in re.finditer(self.variable_ref_patt, rand_cnt, flags=re.DOTALL):
             if self.check_if_canceled("Message streaming"):
                 return
@@ -149,11 +207,9 @@ class Message(ComponentBase):
                 continue
             elif inspect.isawaitable(v):
                 v = await v
-            elif not isinstance(v, str):
-                try:
-                    v = json.dumps(v, ensure_ascii=False)
-                except Exception:
-                    v = str(v)
+            v = self._stringify_message_value(
+                v, downloads=downloads, fallback_to_str=True
+            )
             yield v
             self.set_input_value(exp, v)
             all_content += v
@@ -166,6 +222,7 @@ class Message(ComponentBase):
             all_content += rand_cnt[s: ]
             yield rand_cnt[s: ]
 
+        self.set_output("downloads", downloads)
         self.set_output("content", all_content)
         self._convert_content(all_content)
         await self._save_to_memory(all_content)
@@ -186,12 +243,14 @@ class Message(ComponentBase):
             self.set_output("content", partial(self._stream, rand_cnt))
             return
 
-        rand_cnt, kwargs = self.get_kwargs(rand_cnt, kwargs)
-        template = Jinja2Template(rand_cnt)
+        downloads = []
+        rand_cnt, kwargs = self.get_kwargs(rand_cnt, kwargs, downloads=downloads)
+        template = _jinja2_sandbox.from_string(rand_cnt)
         try:
             content = template.render(kwargs)
-        except Exception:
-            pass
+        except Exception as e:
+            logging.warning(f"Jinja2 template rendering failed: {e}")
+            content = rand_cnt  # fallback to unrendered content
 
         if self.check_if_canceled("Message processing"):
             return
@@ -199,6 +258,7 @@ class Message(ComponentBase):
         for n, v in kwargs.items():
             content = re.sub(n, v, content)
 
+        self.set_output("downloads", downloads)
         self.set_output("content", content)
         self._convert_content(content)
         self._save_to_memory(content)
@@ -224,6 +284,38 @@ class Message(ComponentBase):
         
         rows = []
         headers = None
+
+        def _coerce_excel_cell_type(cell: str):
+            # Convert markdown cell text to native numeric types when safe,so Excel writes numeric cells instead of text.
+            if not isinstance(cell, str):
+                return cell
+
+            value = cell.strip()
+            if value == "":
+                return ""
+
+            # Keep values like "00123" as text to avoid losing leading zeros.
+            if re.match(r"^[+-]?0\d+$", value):
+                return cell
+
+            # Support thousand separators like 1,234 or 1,234.56
+            numeric_candidate = value
+            if re.match(r"^[+-]?\d{1,3}(,\d{3})+(\.\d+)?$", value):
+                numeric_candidate = value.replace(",", "")
+
+            if re.match(r"^[+-]?\d+$", numeric_candidate):
+                try:
+                    return int(numeric_candidate)
+                except ValueError:
+                    return cell
+
+            if re.match(r"^[+-]?(\d+\.\d+|\d+\.|\.\d+)([eE][+-]?\d+)?$", numeric_candidate) or re.match(r"^[+-]?\d+[eE][+-]?\d+$", numeric_candidate):
+                try:
+                    return float(numeric_candidate)
+                except ValueError:
+                    return cell
+
+            return cell
         
         for line in table_lines:
             # Split by | and clean up
@@ -234,6 +326,7 @@ class Message(ComponentBase):
             if headers is None:
                 headers = cells
             else:
+                cells = [_coerce_excel_cell_type(c) for c in cells]
                 rows.append(cells)
         
         if headers and rows:
@@ -430,8 +523,15 @@ class Message(ComponentBase):
         if not hasattr(self._param, "memory_ids") or not self._param.memory_ids:
             return True, "No memory selected."
 
+        user_id = self._param.user_id if hasattr(self._param, "user_id") else ""
+        if user_id:
+            import re
+            # is variable
+            if re.match(r"^{.*}$", user_id):
+                user_id = self._canvas.get_variable_value(user_id)
+
         message_dict = {
-            "user_id": self._canvas._tenant_id,
+            "user_id": user_id,
             "agent_id": self._canvas._id,
             "session_id": self._canvas.task_id,
             "user_input": self._canvas.get_sys_query(),

agent/component/string_transform.py

@@ -18,7 +18,9 @@ import re
 from abc import ABC
 from typing import Any
 
-from jinja2 import Template as Jinja2Template
+from jinja2.sandbox import SandboxedEnvironment
+
+_jinja2_sandbox = SandboxedEnvironment()
 from agent.component.base import ComponentParamBase
 from common.connection_utils import timeout
 from .message import Message
@@ -96,7 +98,7 @@ class StringTransform(Message, ABC):
         script, kwargs = self.get_kwargs(script, kwargs, self._param.delimiters[0])
 
         if self._is_jinjia2(script):
-            template = Jinja2Template(script)
+            template = _jinja2_sandbox.from_string(script)
             try:
                 script = template.render(kwargs)
             except Exception:

agent/component/switch.py

@@ -134,7 +134,7 @@ class Switch(ComponentBase, ABC):
             except Exception:
                 return True if input <= value else False
 
-        raise ValueError('Not supported operator' + operator)
+        raise ValueError(f'Not supported operator: {operator}')
 
     def thoughts(self) -> str:
         return "I’m weighing a few options and will pick the next step shortly."

agent/dsl_migration.py

@@ -0,0 +1,178 @@
+#
+#  Copyright 2026 The InfiniFlow Authors. All Rights Reserved.
+#
+#  Licensed under the Apache License, Version 2.0 (the "License");
+#  you may not use this file except in compliance with the License.
+#  You may obtain a copy of the License at
+#
+#      http://www.apache.org/licenses/LICENSE-2.0
+#
+#  Unless required by applicable law or agreed to in writing, software
+#  distributed under the License is distributed on an "AS IS" BASIS,
+#  WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+#  See the License for the specific language governing permissions and
+#  limitations under the License.
+#
+
+import copy
+import re
+
+
+# Keep all legacy chunker renames in one place so the migration rule stays readable.
+COMPONENT_RENAMES = {
+    "Splitter": "TokenChunker",
+    "HierarchicalMerger": "TitleChunker",
+    "PDFGenerator": "DocGenerator",
+}
+
+NODE_TYPE_RENAMES = {
+    "splitterNode": "chunkerNode",
+}
+
+VARIABLE_REF_PATTERN = re.compile(r"(\{+\s*)([A-Za-z0-9:_-]+)(@[A-Za-z0-9_.-]+)(\s*\}+)")
+
+
+def normalize_chunker_dsl(dsl: dict) -> dict:
+    """
+    Rewrite legacy chunker component names and ids into the current DSL schema.
+
+    This is intentionally a pure migration step:
+    - it does not change business params
+    - it only rewrites structural identifiers used by the canvas/runtime
+    - custom human-authored names are preserved unless they are still the exact
+      built-in legacy operator name
+    """
+    if not isinstance(dsl, dict):
+        return dsl
+
+    normalized = copy.deepcopy(dsl)
+    components = normalized.get("components")
+    if not isinstance(components, dict):
+        return normalized
+
+    component_id_map: dict[str, str] = {}
+    for component_id in components.keys():
+        new_component_id = component_id
+        for old_name, new_name in COMPONENT_RENAMES.items():
+            prefix = f"{old_name}:"
+            if component_id.startswith(prefix):
+                new_component_id = f"{new_name}:{component_id[len(prefix):]}"
+                break
+        component_id_map[component_id] = new_component_id
+
+    def rewrite_variable_refs(text: str) -> str:
+        if text in component_id_map:
+            return component_id_map[text]
+
+        def repl(match: re.Match[str]) -> str:
+            component_id = match.group(2)
+            return (
+                match.group(1)
+                + component_id_map.get(component_id, component_id)
+                + match.group(3)
+                + match.group(4)
+            )
+
+        return VARIABLE_REF_PATTERN.sub(repl, text)
+
+    def rewrite_value(value):
+        if isinstance(value, str):
+            return rewrite_variable_refs(value)
+        if isinstance(value, list):
+            return [rewrite_value(item) for item in value]
+        if isinstance(value, dict):
+            return {key: rewrite_value(item) for key, item in value.items()}
+        return value
+
+    rewritten_components = {}
+    for old_component_id, component in components.items():
+        new_component_id = component_id_map[old_component_id]
+        new_component = rewrite_value(component)
+
+        if isinstance(new_component, dict):
+            obj = new_component.get("obj")
+            if isinstance(obj, dict):
+                component_name = obj.get("component_name")
+                obj["component_name"] = COMPONENT_RENAMES.get(component_name, component_name)
+
+            if isinstance(new_component.get("downstream"), list):
+                new_component["downstream"] = [
+                    component_id_map.get(component_id, component_id)
+                    for component_id in new_component["downstream"]
+                ]
+            if isinstance(new_component.get("upstream"), list):
+                new_component["upstream"] = [
+                    component_id_map.get(component_id, component_id)
+                    for component_id in new_component["upstream"]
+                ]
+
+            parent_id = new_component.get("parent_id")
+            if isinstance(parent_id, str):
+                new_component["parent_id"] = component_id_map.get(parent_id, parent_id)
+
+        rewritten_components[new_component_id] = new_component
+
+    normalized["components"] = rewritten_components
+
+    if isinstance(normalized.get("path"), list):
+        normalized["path"] = [
+            component_id_map.get(component_id, component_id)
+            for component_id in normalized["path"]
+        ]
+
+    graph = normalized.get("graph")
+    if isinstance(graph, dict):
+        nodes = graph.get("nodes")
+        if isinstance(nodes, list):
+            for node in nodes:
+                if not isinstance(node, dict):
+                    continue
+                node_id = node.get("id")
+                if isinstance(node_id, str):
+                    node["id"] = component_id_map.get(node_id, node_id)
+
+                parent_id = node.get("parentId")
+                if isinstance(parent_id, str):
+                    node["parentId"] = component_id_map.get(parent_id, parent_id)
+
+                node_type = node.get("type")
+                if isinstance(node_type, str):
+                    node["type"] = NODE_TYPE_RENAMES.get(node_type, node_type)
+
+                data = node.get("data")
+                if not isinstance(data, dict):
+                    continue
+
+                label = data.get("label")
+                if isinstance(label, str):
+                    data["label"] = COMPONENT_RENAMES.get(label, label)
+
+                name = data.get("name")
+                if isinstance(name, str) and name in COMPONENT_RENAMES:
+                    data["name"] = COMPONENT_RENAMES[name]
+
+                if "form" in data:
+                    data["form"] = rewrite_value(data["form"])
+
+        edges = graph.get("edges")
+        if isinstance(edges, list):
+            replacements = sorted(component_id_map.items(), key=lambda item: len(item[0]), reverse=True)
+            for edge in edges:
+                if not isinstance(edge, dict):
+                    continue
+                for key in ("source", "target"):
+                    value = edge.get(key)
+                    if isinstance(value, str):
+                        edge[key] = component_id_map.get(value, value)
+
+                edge_id = edge.get("id")
+                if isinstance(edge_id, str):
+                    for old_component_id, new_component_id in replacements:
+                        edge_id = edge_id.replace(old_component_id, new_component_id)
+                    edge["id"] = edge_id
+
+    for key in ("history", "messages", "reference"):
+        if key in normalized:
+            normalized[key] = rewrite_value(normalized[key])
+
+    return normalized

agent/plugin/README_tr.md

@@ -0,0 +1,99 @@
+[English](./README.md) | [简体中文](./README_zh.md) | Türkçe
+
+# Eklentiler
+
+Bu klasör, RAGFlow'un eklenti mekanizmasını içerir.
+
+RAGFlow, `embedded_plugins` alt klasöründen eklentileri özyinelemeli olarak yükleyecektir.
+
+## Desteklenen eklenti türleri
+
+Şu anda desteklenen tek eklenti türü `llm_tools`'dur.
+
+- `llm_tools`: LLM'nin çağırması için bir araç.
+
+## Eklenti nasıl eklenir
+
+Bir LLM araç eklentisi eklemek basittir: bir eklenti dosyası oluşturun, içine `LLMToolPlugin` sınıfından türetilmiş bir sınıf koyun, ardından `get_metadata` ve `invoke` metodlarını uygulayın.
+
+- `get_metadata` metodu: Bu metod, aracın açıklamasını içeren bir `LLMToolMetadata` nesnesi döndürür.
+Açıklama, LLM'ye çağrı için ve RAGFlow web ön yüzüne görüntüleme amacıyla sağlanacaktır.
+
+- `invoke` metodu: Bu metod, LLM tarafından üretilen parametreleri kabul eder ve aracın yürütme sonucunu içeren bir `str` döndürür.
+Bu aracın tüm yürütme mantığı bu metoda konulmalıdır.
+
+RAGFlow'u başlattığınızda, günlükte eklentinizin yüklendiğini göreceksiniz:
+
+```
+2025-05-15 19:29:08,959 INFO     34670 Recursively importing plugins from path `/some-path/ragflow/agent/plugin/embedded_plugins`
+2025-05-15 19:29:08,960 INFO     34670 Loaded llm_tools plugin BadCalculatorPlugin version 1.0.0
+```
+
+Veya eklentinizi düzeltmeniz gereken hatalar da içerebilir.
+
+### Örnek
+
+Yanlış cevaplar veren bir hesap makinesi aracı ekleyerek eklenti ekleme sürecini göstereceğiz.
+
+Önce, `embedded_plugins/llm_tools` klasörü altında `bad_calculator.py` adında bir eklenti dosyası oluşturun.
+
+Ardından, `LLMToolPlugin` temel sınıfından türetilmiş bir `BadCalculatorPlugin` sınıfı oluşturuyoruz:
+
+```python
+class BadCalculatorPlugin(LLMToolPlugin):
+    _version_ = "1.0.0"
+```
+
+`_version_` alanı zorunludur ve eklentinin sürüm numarasını belirtir.
+
+Hesap makinemizin girdileri olarak `a` ve `b` olmak üzere iki sayısı vardır, bu yüzden `BadCalculatorPlugin` sınıfımıza aşağıdaki `invoke` metodunu ekliyoruz:
+
+```python
+def invoke(self, a: int, b: int) -> str:
+    return str(a + b + 100)
+```
+
+`invoke` metodu LLM tarafından çağrılacaktır. Birçok parametreye sahip olabilir, ancak dönüş tipi `str` olmalıdır.
+
+Son olarak, LLM'ye `bad_calculator` aracımızı nasıl kullanacağını anlatmak için bir `get_metadata` metodu eklememiz gerekiyor:
+
+```python
+@classmethod
+def get_metadata(cls) -> LLMToolMetadata:
+    return {
+        # Bu aracın adı, LLM'ye sağlanır
+        "name": "bad_calculator",
+        # Bu aracın görüntüleme adı, RAGFlow ön yüzüne sağlanır
+        "displayName": "$t:bad_calculator.name",
+        # Bu aracın kullanım açıklaması, LLM'ye sağlanır
+        "description": "A tool to calculate the sum of two numbers (will give wrong answer)",
+        # Bu aracın açıklaması, RAGFlow ön yüzüne sağlanır
+        "displayDescription": "$t:bad_calculator.description",
+        # Bu aracın parametreleri
+        "parameters": {
+            # Birinci parametre - a
+            "a": {
+                # Parametre tipi, seçenekler: number, string veya LLM'nin tanıyabileceği herhangi bir tip
+                "type": "number",
+                # Bu parametrenin açıklaması, LLM'ye sağlanır
+                "description": "The first number",
+                # Bu parametrenin açıklaması, RAGFlow ön yüzüne sağlanır
+                "displayDescription": "$t:bad_calculator.params.a",
+                # Bu parametrenin zorunlu olup olmadığı
+                "required": True
+            },
+            # İkinci parametre - b
+            "b": {
+                "type": "number",
+                "description": "The second number",
+                "displayDescription": "$t:bad_calculator.params.b",
+                "required": True
+            }
+        }
+```
+
+`get_metadata` metodu bir `classmethod`'dur. Bu aracın açıklamasını LLM'ye sağlayacaktır.
+
+`display` ile başlayan alanlar özel bir gösterim kullanabilir: `$t:xxx`, bu gösterim RAGFlow ön yüzündeki uluslararasılaştırma (i18n) mekanizmasını kullanarak `llmTools` kategorisinden metin alır. Bu gösterimi kullanmazsanız, ön yüz buraya yazdığınız metni doğrudan gösterecektir.
+
+Artık aracımız hazırdır. `Yanıt Üret` bileşeninde seçip deneyebilirsiniz.

agent/sandbox/README.md

@@ -189,7 +189,19 @@ Currently, the following languages are officially supported:
 
 ### 🐍 Python
 
-To add Python dependencies, simply edit the following file:
+Pre-installed packages: `requests`, `numpy`, `pandas`, `matplotlib`.
+
+> `matplotlib` uses the `Agg` (non-interactive) backend by default in the sandbox (`MPLBACKEND=Agg`). No display server is available, so always save figures to files (e.g. `fig.savefig("artifacts/chart.png")`) rather than calling `plt.show()`.
+>
+> Tip: if Chinese text renders as missing boxes/squares in `matplotlib`, install Debian package `fonts-noto-cjk` in your custom image. We do not preinstall it by default to keep the base image smaller. The sandbox base image ships a `matplotlibrc` that already lists common CJK fonts in the `font.sans-serif` fallback chain, so no code-level font configuration is needed — just install the font package and rebuild the image.
+>
+> Example:
+>
+> ```dockerfile
+> RUN apt-get update && apt-get install -y --no-install-recommends fonts-noto-cjk && rm -rf /var/lib/apt/lists/*
+> ```
+
+To add more dependencies, edit:
 
 ```bash
 sandbox_base_image/python/requirements.txt
@@ -199,6 +211,8 @@ Add any additional packages you need, one per line (just like a normal pip requi
 
 ### 🟨 Node.js
 
+Pre-installed packages: `axios`.
+
 To add Node.js dependencies:
 
 1. Navigate to the Node.js base image directory:

agent/sandbox/docker-compose.yml

@@ -7,7 +7,7 @@ services:
     runtime: runc
     privileged: true
     ports:
-      - "${EXECUTOR_PORT:-9385}:9385"
+      - "${SANDBOX_EXECUTOR_MANAGER_PORT:-9385}:9385"
     volumes:
       - /var/run/docker.sock:/var/run/docker.sock
     networks:

agent/sandbox/executor_manager/api/routes.py

@@ -19,6 +19,7 @@ from api.handlers import healthz_handler, run_code_handler
 
 router = APIRouter()
 
+router.get("/")(healthz_handler)
 router.get("/healthz")(healthz_handler)
 router.post("/run")(run_code_handler)
 

agent/sandbox/executor_manager/models/schemas.py

@@ -14,13 +14,26 @@
 #  limitations under the License.
 #
 import base64
-from typing import Optional
+from typing import Any, Optional
 
 from pydantic import BaseModel, Field, field_validator
 
 from models.enums import ResourceLimitType, ResultStatus, RuntimeErrorType, SupportLanguage, UnauthorizedAccessType
 
 
+class ArtifactItem(BaseModel):
+    name: str
+    mime_type: str
+    size: int
+    content_b64: str
+
+
+class ExecutionStructuredResult(BaseModel):
+    present: bool
+    value: Any = None
+    type: str = "json"
+
+
 class CodeExecutionResult(BaseModel):
     status: ResultStatus
     stdout: str
@@ -37,6 +50,12 @@ class CodeExecutionResult(BaseModel):
     unauthorized_access_type: Optional[UnauthorizedAccessType] = None
     runtime_error_type: Optional[RuntimeErrorType] = None
 
+    # File artifacts produced by code execution (images, PDFs, CSVs, etc.)
+    artifacts: list[ArtifactItem] = []
+
+    # Structured return value produced by main()
+    result: Optional[ExecutionStructuredResult] = None
+
 
 class CodeExecutionRequest(BaseModel):
     code_b64: str = Field(..., description="Base64 encoded code string")

agent/sandbox/executor_manager/services/execution.py

@@ -19,17 +19,181 @@ import json
 import os
 import time
 import uuid
-
 from core.config import TIMEOUT
 from core.container import allocate_container_blocking, release_container
 from core.logger import logger
 from models.enums import ResourceLimitType, ResultStatus, RuntimeErrorType, SupportLanguage, UnauthorizedAccessType
-from models.schemas import CodeExecutionRequest, CodeExecutionResult
+from models.schemas import ArtifactItem, CodeExecutionRequest, CodeExecutionResult, ExecutionStructuredResult
 from utils.common import async_run_command
 
+RESULT_MARKER_PREFIX = "__RAGFLOW_RESULT__:"
+
+
+def _extract_result_envelope(stdout: str) -> tuple[str, ExecutionStructuredResult | None]:
+    if not stdout:
+        return "", None
+
+    cleaned_lines: list[str] = []
+    envelope: ExecutionStructuredResult | None = None
+
+    for line in str(stdout).splitlines():
+        if line.startswith(RESULT_MARKER_PREFIX):
+            payload_b64 = line[len(RESULT_MARKER_PREFIX) :].strip()
+            if not payload_b64:
+                continue
+            try:
+                payload = base64.b64decode(payload_b64).decode("utf-8")
+                envelope = ExecutionStructuredResult.model_validate_json(payload)
+            except Exception as exc:
+                logger.warning(f"Failed to decode structured result marker: {exc}")
+                cleaned_lines.append(line)
+            continue
+        cleaned_lines.append(line)
+
+    cleaned_stdout = "\n".join(cleaned_lines)
+    if stdout.endswith("\n") and cleaned_stdout and not cleaned_stdout.endswith("\n"):
+        cleaned_stdout += "\n"
+    return cleaned_stdout, envelope
+
+
+def _build_execution_bundle(req: CodeExecutionRequest, workdir: str) -> dict[str, str | bytes]:
+    arguments = req.arguments or {}
+    args_source = json.dumps(arguments, ensure_ascii=False)
+    args_name = "args.json"
+    code_bytes = base64.b64decode(req.code_b64)
+
+    if req.language == SupportLanguage.PYTHON:
+        code_name = "main.py"
+        runner_name = "runner.py"
+        runner_source = f"""import base64
+import json
+import os
+import sys
+
+os.makedirs(os.path.join(os.getcwd(), "artifacts"), exist_ok=True)
+
+sys.path.insert(0, os.path.dirname(__file__))
+from main import main
+
+RESULT_MARKER_PREFIX = {RESULT_MARKER_PREFIX!r}
+
+
+def emit_result(value):
+    payload = json.dumps(
+        {{
+            "present": True,
+            "value": value,
+            "type": "json",
+        }},
+        ensure_ascii=False,
+        separators=(",", ":"),
+    )
+    print(RESULT_MARKER_PREFIX + base64.b64encode(payload.encode("utf-8")).decode("ascii"))
+
+
+if __name__ == "__main__":
+    with open(os.path.join(os.path.dirname(__file__), "args.json"), encoding="utf-8") as f:
+        args = json.load(f)
+    result = main(**args)
+    emit_result(result)
+"""
+    elif req.language == SupportLanguage.NODEJS:
+        code_name = "main.js"
+        runner_name = "runner.js"
+        runner_source = """
+const fs = require('fs');
+const path = require('path');
+
+const args = JSON.parse(fs.readFileSync(path.join(__dirname, 'args.json'), 'utf8'));
+const mainPath = path.join(__dirname, 'main.js');
+const RESULT_MARKER_PREFIX = '__RESULT_MARKER_PREFIX__';
+
+function isPromise(value) {
+    return Boolean(value && typeof value.then === 'function');
+}
+
+function emitResult(value) {
+    if (typeof value === 'undefined') {
+        console.error('Error: main() must return a value. Use null for an empty result.');
+        process.exit(1);
+    }
+
+    const payload = JSON.stringify({ present: true, value, type: 'json' });
+    if (typeof payload === 'undefined') {
+        console.error('Error: main() returned a non-JSON-serializable value.');
+        process.exit(1);
+    }
+
+    console.log(RESULT_MARKER_PREFIX + Buffer.from(payload, 'utf8').toString('base64'));
+}
+
+if (fs.existsSync(mainPath)) {
+    const mod = require(mainPath);
+    const main = typeof mod === 'function' ? mod : mod.main;
+
+    if (typeof main !== 'function') {
+        console.error('Error: main is not a function');
+        process.exit(1);
+    }
+
+    if (typeof args === 'object' && args !== null) {
+        try {
+            const result = Promise.resolve(main(args));
+            if (isPromise(result)) {
+                result.then(output => {
+                    emitResult(output);
+                }).catch(err => {
+                    console.error('Error in async main function:', err);
+                    process.exit(1);
+                });
+            } else {
+                emitResult(result);
+            }
+        } catch (err) {
+            console.error('Error when executing main:', err);
+            process.exit(1);
+        }
+    } else {
+        console.error('Error: args is not a valid object:', args);
+        process.exit(1);
+    }
+} else {
+    console.error('main.js not found in the current directory');
+    process.exit(1);
+}
+"""
+        runner_source = runner_source.replace("__RESULT_MARKER_PREFIX__", RESULT_MARKER_PREFIX)
+    else:
+        assert False, "Will never reach here"
+
+    return {
+        "code_name": code_name,
+        "code_bytes": code_bytes,
+        "runner_name": runner_name,
+        "runner_source": runner_source,
+        "args_name": args_name,
+        "args_source": args_source,
+    }
+
+
+def _build_container_run_args(language: SupportLanguage, task_id: str, container: str, runner_name: str) -> list[str]:
+    run_args = [
+        "docker",
+        "exec",
+        "--workdir",
+        f"/workspace/{task_id}",
+        container,
+        "timeout",
+        str(TIMEOUT),
+        language,
+    ]
+    if language == SupportLanguage.PYTHON:
+        run_args.extend(["-I", "-B"])
+    run_args.append(runner_name)
+    return run_args
+
 
 async def execute_code(req: CodeExecutionRequest):
-    """Fully asynchronous execution logic"""
     language = req.language
     container = await allocate_container_blocking(language)
     if not container:
@@ -46,93 +210,31 @@ async def execute_code(req: CodeExecutionRequest):
     os.makedirs(workdir, mode=0o700, exist_ok=True)
 
     try:
-        if language == SupportLanguage.PYTHON:
-            code_name = "main.py"
-            # code
-            code_path = os.path.join(workdir, code_name)
-            with open(code_path, "wb") as f:
-                f.write(base64.b64decode(req.code_b64))
-            # runner
-            runner_name = "runner.py"
-            runner_path = os.path.join(workdir, runner_name)
-            with open(runner_path, "w") as f:
-                f.write("""import json
-import os
-import sys
-sys.path.insert(0, os.path.dirname(__file__))
-from main import main
-if __name__ == "__main__":
-    args = json.loads(sys.argv[1])
-    result = main(**args)
-    if result is not None:
-        print(result)
-""")
+        bundle = _build_execution_bundle(req, workdir)
+        code_name = str(bundle["code_name"])
+        runner_name = str(bundle["runner_name"])
 
-        elif language == SupportLanguage.NODEJS:
-            code_name = "main.js"
-            code_path = os.path.join(workdir, "main.js")
-            with open(code_path, "wb") as f:
-                f.write(base64.b64decode(req.code_b64))
+        code_path = os.path.join(workdir, code_name)
+        with open(code_path, "wb") as f:
+            f.write(bundle["code_bytes"])
 
-            runner_name = "runner.js"
-            runner_path = os.path.join(workdir, "runner.js")
-            with open(runner_path, "w") as f:
-                f.write("""
-const fs = require('fs');
-const path = require('path');
+        runner_path = os.path.join(workdir, runner_name)
+        with open(runner_path, "w", encoding="utf-8") as f:
+            f.write(str(bundle["runner_source"]))
 
-const args = JSON.parse(process.argv[2]);
-const mainPath = path.join(__dirname, 'main.js');
+        args_path = os.path.join(workdir, str(bundle["args_name"]))
+        with open(args_path, "w", encoding="utf-8") as f:
+            f.write(str(bundle["args_source"]))
 
-function isPromise(value) {
-    return Boolean(value && typeof value.then === 'function');
-}
-
-if (fs.existsSync(mainPath)) {
-    const mod = require(mainPath);
-    const main = typeof mod === 'function' ? mod : mod.main;
-
-    if (typeof main !== 'function') {
-        console.error('Error: main is not a function');
-        process.exit(1);
-    }
-
-    if (typeof args === 'object' && args !== null) {
-        try {
-            const result = main(args);
-            if (isPromise(result)) {
-                result.then(output => {
-                    if (output !== null) {
-                        console.log(output);
-                    }
-                }).catch(err => {
-                    console.error('Error in async main function:', err);
-                });
-            } else {
-                if (result !== null) {
-                    console.log(result);
-                }
-            }
-        } catch (err) {
-            console.error('Error when executing main:', err);
-        }
-    } else {
-        console.error('Error: args is not a valid object:', args);
-    }
-} else {
-    console.error('main.js not found in the current directory');
-}
-""")
-        # dirs
         returncode, _, stderr = await async_run_command("docker", "exec", container, "mkdir", "-p", f"/workspace/{task_id}", timeout=5)
         if returncode != 0:
             raise RuntimeError(f"Directory creation failed: {stderr}")
 
-        # archive
-        tar_proc = await asyncio.create_subprocess_exec("tar", "czf", "-", "-C", workdir, code_name, runner_name, stdout=asyncio.subprocess.PIPE)
+        tar_proc = await asyncio.create_subprocess_exec(
+            "tar", "czf", "-", "-C", workdir, code_name, runner_name, str(bundle["args_name"]), stdout=asyncio.subprocess.PIPE
+        )
         tar_stdout, _ = await tar_proc.communicate()
 
-        # unarchive
         docker_proc = await asyncio.create_subprocess_exec(
             "docker", "exec", "-i", container, "tar", "xzf", "-", "-C", f"/workspace/{task_id}", stdin=asyncio.subprocess.PIPE, stderr=asyncio.subprocess.PIPE
         )
@@ -141,29 +243,11 @@ if (fs.existsSync(mainPath)) {
         if docker_proc.returncode != 0:
             raise RuntimeError(stderr.decode())
 
-        # exec
         start_time = time.time()
         try:
-            logger.info(f"Passed in args: {req.arguments}")
-            args_json = json.dumps(req.arguments or {})
-            run_args = [
-                "docker",
-                "exec",
-                "--workdir",
-                f"/workspace/{task_id}",
-                container,
-                "timeout",
-                str(TIMEOUT),
-                language,
-            ]
-            # flags
-            if language == SupportLanguage.PYTHON:
-                run_args.extend(["-I", "-B"])
-            elif language == SupportLanguage.NODEJS:
-                run_args.extend([])
-            else:
-                assert False, "Will never reach here"
-            run_args.extend([runner_name, args_json])
+            arguments = req.arguments or {}
+            logger.info("Passed in args keys=%s size_bytes=%s", list(arguments.keys()), len(json.dumps(arguments, ensure_ascii=False).encode("utf-8")))
+            run_args = _build_container_run_args(language=language, task_id=task_id, container=container, runner_name=runner_name)
 
             returncode, stdout, stderr = await async_run_command(
                 *run_args,
@@ -177,15 +261,18 @@ if (fs.existsSync(mainPath)) {
             logger.info(f"{returncode=}")
             logger.info(f"{stdout=}")
             logger.info(f"{stderr=}")
-            logger.info(f"{args_json=}")
 
             if returncode == 0:
+                clean_stdout, structured_result = _extract_result_envelope(stdout)
+                artifacts = await _collect_artifacts(container, task_id, workdir)
                 return CodeExecutionResult(
                     status=ResultStatus.SUCCESS,
-                    stdout=str(stdout),
+                    stdout=clean_stdout,
                     stderr=stderr,
                     exit_code=0,
                     time_used_ms=time_used_ms,
+                    artifacts=artifacts,
+                    result=structured_result,
                 )
             elif returncode == 124:
                 return CodeExecutionResult(
@@ -223,12 +310,89 @@ if (fs.existsSync(mainPath)) {
         return CodeExecutionResult(status=ResultStatus.PROGRAM_RUNNER_ERROR, stdout="", stderr=str(e), exit_code=-3, detail="internal_error")
 
     finally:
-        # cleanup
         cleanup_tasks = [async_run_command("docker", "exec", container, "rm", "-rf", f"/workspace/{task_id}"), async_run_command("rm", "-rf", workdir)]
         await asyncio.gather(*cleanup_tasks, return_exceptions=True)
         await release_container(container, language)
 
 
+ALLOWED_ARTIFACT_EXTENSIONS = {
+    ".png": "image/png",
+    ".jpg": "image/jpeg",
+    ".jpeg": "image/jpeg",
+    ".svg": "image/svg+xml",
+    ".pdf": "application/pdf",
+    ".csv": "text/csv",
+    ".json": "application/json",
+    ".html": "text/html",
+}
+MAX_ARTIFACT_COUNT = 10
+MAX_ARTIFACT_SIZE = 10 * 1024 * 1024  # 10MB per file
+
+
+async def _collect_artifacts(container: str, task_id: str, host_workdir: str) -> list[ArtifactItem]:
+    artifacts_path = f"/workspace/{task_id}/artifacts"
+
+    # List files in the artifacts directory inside the container
+    returncode, stdout, _ = await async_run_command(
+        "docker", "exec", container, "find", artifacts_path,
+        "-maxdepth", "1", "-type", "f", timeout=5,
+    )
+    if returncode != 0 or not stdout.strip():
+        return []
+
+    raw_names = [line.split("/")[-1] for line in stdout.strip().splitlines() if line.strip()]
+    # Sanitize: reject names with path traversal or control characters
+    filenames = [n for n in raw_names if n and "/" not in n and "\\" not in n and ".." not in n and not n.startswith(".")]
+    if not filenames:
+        return []
+
+    items: list[ArtifactItem] = []
+
+    for fname in filenames[:MAX_ARTIFACT_COUNT]:
+        ext = os.path.splitext(fname)[1].lower()
+        mime_type = ALLOWED_ARTIFACT_EXTENSIONS.get(ext)
+        if not mime_type:
+            logger.warning(f"Skipping artifact with disallowed extension: {fname}")
+            continue
+
+        file_path = f"{artifacts_path}/{fname}"
+
+        # Check file size inside the container
+        returncode, size_str, _ = await async_run_command(
+            "docker", "exec", container, "stat", "-c", "%s", file_path, timeout=5,
+        )
+        if returncode != 0:
+            logger.warning(f"Failed to stat artifact {fname}")
+            continue
+
+        file_size = int(size_str.strip())
+        if file_size > MAX_ARTIFACT_SIZE:
+            logger.warning(f"Artifact {fname} too large ({file_size} bytes), skipping")
+            continue
+        if file_size == 0:
+            continue
+
+        # Read file content via docker exec (docker cp doesn't work with gVisor tmpfs)
+        returncode, content_b64, stderr = await async_run_command(
+            "docker", "exec", container, "base64", file_path, timeout=30,
+        )
+        if returncode != 0:
+            logger.warning(f"Failed to read artifact {fname}: {stderr}")
+            continue
+
+        content_b64 = content_b64.replace("\n", "").strip()
+
+        items.append(ArtifactItem(
+            name=fname,
+            mime_type=mime_type,
+            size=file_size,
+            content_b64=content_b64,
+        ))
+        logger.info(f"Collected artifact: {fname} ({file_size} bytes, {mime_type})")
+
+    return items
+
+
 def analyze_error_result(stderr: str, exit_code: int) -> CodeExecutionResult:
     """Analyze the error result and classify it"""
     if "Permission denied" in stderr:

agent/sandbox/executor_manager/services/security.py

@@ -14,6 +14,7 @@
 #  limitations under the License.
 #
 import ast
+import re
 from typing import List, Tuple
 
 from core.logger import logger
@@ -151,6 +152,26 @@ class SecurePythonAnalyzer(ast.NodeVisitor):
         self.generic_visit(node)
 
 
+class SecureJavaScriptAnalyzer:
+    DANGEROUS_PATTERNS = [
+        (re.compile(r"""require\s*\(\s*['"]child_process['"]\s*\)"""), "Require: child_process"),
+        (re.compile(r"""require\s*\(\s*['"]fs['"]\s*\)"""), "Require: fs"),
+        (re.compile(r"""require\s*\(\s*['"]worker_threads['"]\s*\)"""), "Require: worker_threads"),
+        (re.compile(r"""\beval\s*\("""), "Call: eval"),
+        (re.compile(r"""\bFunction\s*\("""), "Call: Function"),
+        (re.compile(r"""\bprocess\s*\.\s*binding\s*\("""), "Call: process.binding"),
+    ]
+
+    @classmethod
+    def analyze(cls, code: str) -> List[Tuple[str, int]]:
+        issues: List[Tuple[str, int]] = []
+        for pattern, description in cls.DANGEROUS_PATTERNS:
+            for match in pattern.finditer(code):
+                lineno = code.count("\n", 0, match.start()) + 1
+                issues.append((description, lineno))
+        return issues
+
+
 def analyze_code_security(code: str, language: SupportLanguage) -> Tuple[bool, List[Tuple[str, int]]]:
     """
     Analyze the provided code string and return whether it's safe and why.
@@ -168,6 +189,9 @@ def analyze_code_security(code: str, language: SupportLanguage) -> Tuple[bool, L
         except Exception as e:
             logger.error(f"[SafeCheck] Python parsing failed: {str(e)}")
             return False, [(f"Parsing Error: {str(e)}", -1)]
-    else:
-        logger.warning(f"[SafeCheck] Unsupported language for security analysis: {language} — defaulting to SAFE (manual review recommended)")
-        return True, [(f"Unsupported language for security analysis: {language} — defaulted to SAFE, manual review recommended", -1)]
+    if language == SupportLanguage.NODEJS:
+        issues = SecureJavaScriptAnalyzer.analyze(code)
+        return len(issues) == 0, issues
+
+    logger.warning(f"[SafeCheck] Unsupported language for security analysis: {language}")
+    return False, [(f"Unsupported language for security analysis: {language}", -1)]

agent/sandbox/providers/aliyun_codeinterpreter.py

@@ -30,6 +30,8 @@ https://api.aliyun.com/api/AgentRun/2025-09-10/CreateSandbox?lang=PYTHON
 import logging
 import os
 import time
+import base64
+import json
 from typing import Dict, Any, List, Optional
 from datetime import datetime, timezone
 
@@ -40,6 +42,7 @@ from agentrun.utils.exception import ServerError
 from .base import SandboxProvider, SandboxInstance, ExecutionResult
 
 logger = logging.getLogger(__name__)
+RESULT_MARKER_PREFIX = "__RAGFLOW_RESULT__:"
 
 
 class AliyunCodeInterpreterProvider(SandboxProvider):
@@ -51,9 +54,9 @@ class AliyunCodeInterpreterProvider(SandboxProvider):
     """
 
     def __init__(self):
-        self.access_key_id: Optional[str] = None
-        self.access_key_secret: Optional[str] = None
-        self.account_id: Optional[str] = None
+        self.access_key_id: Optional[str] = ""
+        self.access_key_secret: Optional[str] = ""
+        self.account_id: Optional[str] = ""
         self.region: str = "cn-hangzhou"
         self.template_name: str = ""
         self.timeout: int = 30
@@ -68,7 +71,7 @@ class AliyunCodeInterpreterProvider(SandboxProvider):
             config: Configuration dictionary with keys:
                 - access_key_id: Aliyun AccessKey ID
                 - access_key_secret: Aliyun AccessKey Secret
-                - account_id: Aliyun primary account ID (主账号ID)
+                - account_id: Aliyun primary account ID
                 - region: Region (default: "cn-hangzhou")
                 - template_name: Optional sandbox template name
                 - timeout: Request timeout in seconds (default: 30, max 30)
@@ -97,7 +100,7 @@ class AliyunCodeInterpreterProvider(SandboxProvider):
             return False
 
         if not self.account_id:
-            logger.error("Aliyun Code Interpreter: Missing account_id (主账号ID)")
+            logger.error("Aliyun Code Interpreter: Missing account_id (primary account ID)")
             return False
 
         # Create SDK configuration
@@ -146,8 +149,6 @@ class AliyunCodeInterpreterProvider(SandboxProvider):
 
         try:
             # Get or create template
-            from agentrun.sandbox import Sandbox
-
             if self.template_name:
                 # Use existing template
                 template_name = self.template_name
@@ -226,48 +227,17 @@ class AliyunCodeInterpreterProvider(SandboxProvider):
             # Connect to existing sandbox instance
             sandbox = Sandbox.connect(sandbox_id=instance_id, config=self._config)
 
-            # Convert language string to CodeLanguage enum
-            code_language = CodeLanguage.PYTHON if normalized_lang == "python" else CodeLanguage.JAVASCRIPT
+            # agentrun-sdk 0.0.26 only exposes CodeLanguage.PYTHON; keep JS as string fallback.
+            code_language = CodeLanguage.PYTHON if normalized_lang == "python" else "javascript"
 
             # Wrap code to call main() function
             # Matches self_managed provider behavior: call main(**arguments)
-            if normalized_lang == "python":
-                # Build arguments string for main() call
-                if arguments:
-                    import json as json_module
-                    args_json = json_module.dumps(arguments)
-                    wrapped_code = f'''{code}
-
-if __name__ == "__main__":
-    import json
-    result = main(**{args_json})
-    print(json.dumps(result) if isinstance(result, dict) else result)
-'''
-                else:
-                    wrapped_code = f'''{code}
-
-if __name__ == "__main__":
-    import json
-    result = main()
-    print(json.dumps(result) if isinstance(result, dict) else result)
-'''
-            else:  # javascript
-                if arguments:
-                    import json as json_module
-                    args_json = json_module.dumps(arguments)
-                    wrapped_code = f'''{code}
-
-// Call main and output result
-const result = main({args_json});
-console.log(typeof result === 'object' ? JSON.stringify(result) : String(result));
-'''
-                else:
-                    wrapped_code = f'''{code}
-
-// Call main and output result
-const result = main();
-console.log(typeof result === 'object' ? JSON.stringify(result) : String(result));
-'''
+            args_json = json.dumps(arguments or {})
+            wrapped_code = (
+                self._build_python_wrapper(code, args_json)
+                if normalized_lang == "python"
+                else self._build_javascript_wrapper(code, args_json)
+            )
             logger.debug(f"Aliyun Code Interpreter: Wrapped code (first 200 chars): {wrapped_code[:200]}")
 
             start_time = time.time()
@@ -314,6 +284,7 @@ console.log(typeof result === 'object' ? JSON.stringify(result) : String(result)
 
             stdout = "\n".join(stdout_parts)
             stderr = "\n".join(stderr_parts)
+            stdout, structured_result = self._extract_structured_result(stdout)
 
             logger.info(f"Aliyun Code Interpreter: stdout length={len(stdout)}, stderr length={len(stderr)}, exit_code={exit_code}")
             if stdout:
@@ -331,6 +302,9 @@ console.log(typeof result === 'object' ? JSON.stringify(result) : String(result)
                     "language": normalized_lang,
                     "context_id": result.get("contextId") if isinstance(result, dict) else None,
                     "timeout": timeout,
+                    "result_present": structured_result.get("present", False),
+                    "result_value": structured_result.get("value"),
+                    "result_type": structured_result.get("type"),
                 },
             )
 
@@ -390,6 +364,71 @@ console.log(typeof result === 'object' ? JSON.stringify(result) : String(result)
             # If we get any response (even an error), the service is reachable
             return "connection" not in str(e).lower()
 
+    @staticmethod
+    def _build_python_wrapper(code: str, args_json: str) -> str:
+        marker = RESULT_MARKER_PREFIX
+        return f'''{code}
+
+if __name__ == "__main__":
+    import base64
+    import json
+
+    result = main(**{args_json})
+    payload = json.dumps({{"present": True, "value": result, "type": "json"}}, ensure_ascii=False, separators=(",", ":"))
+    print("{marker}" + base64.b64encode(payload.encode("utf-8")).decode("ascii"))
+'''
+
+    @staticmethod
+    def _build_javascript_wrapper(code: str, args_json: str) -> str:
+        marker = RESULT_MARKER_PREFIX
+        return f'''{code}
+
+const __ragflowArgs = {args_json};
+
+(async () => {{
+  try {{
+    const output = await Promise.resolve(main(__ragflowArgs));
+    if (typeof output === 'undefined') {{
+      throw new Error('main() must return a value. Use null for an empty result.');
+    }}
+    const payload = JSON.stringify({{ present: true, value: output, type: 'json' }});
+    if (typeof payload === 'undefined') {{
+      throw new Error('main() returned a non-JSON-serializable value.');
+    }}
+    console.log('{marker}' + Buffer.from(payload, 'utf8').toString('base64'));
+  }} catch (err) {{
+    console.error(err instanceof Error ? err.stack || err.message : String(err));
+  }}
+}})();
+'''
+
+    @staticmethod
+    def _extract_structured_result(stdout: str) -> tuple[str, Dict[str, Any]]:
+        if not stdout:
+            return "", {}
+
+        cleaned_lines: list[str] = []
+        structured_result: Dict[str, Any] = {}
+
+        for line in str(stdout).splitlines():
+            if line.startswith(RESULT_MARKER_PREFIX):
+                payload_b64 = line[len(RESULT_MARKER_PREFIX) :].strip()
+                if not payload_b64:
+                    continue
+                try:
+                    payload = base64.b64decode(payload_b64).decode("utf-8")
+                    structured_result = json.loads(payload)
+                except Exception as exc:
+                    logger.warning(f"Aliyun Code Interpreter: failed to decode structured result marker: {exc}")
+                    cleaned_lines.append(line)
+                continue
+            cleaned_lines.append(line)
+
+        cleaned_stdout = "\n".join(cleaned_lines)
+        if stdout.endswith("\n") and cleaned_stdout and not cleaned_stdout.endswith("\n"):
+            cleaned_stdout += "\n"
+        return cleaned_stdout, structured_result
+
     def get_supported_languages(self) -> List[str]:
         """
         Get list of supported programming languages.
@@ -429,7 +468,7 @@ console.log(typeof result === 'object' ? JSON.stringify(result) : String(result)
                 "required": True,
                 "label": "Account ID",
                 "placeholder": "1234567890...",
-                "description": "Aliyun primary account ID (主账号ID), required for API calls",
+                "description": "Aliyun primary account ID, required for API calls",
             },
             "region": {
                 "type": "string",

agent/sandbox/providers/self_managed.py

@@ -70,7 +70,7 @@ class SelfManagedProvider(SandboxProvider):
             # Try to fall back to SANDBOX_HOST from settings if we are using localhost
             if "localhost" in self.endpoint or "127.0.0.1" in self.endpoint:
                 try:
-                    from api import settings
+                    from common import settings
                     if settings.SANDBOX_HOST and settings.SANDBOX_HOST not in self.endpoint:
                         original_endpoint = self.endpoint
                         self.endpoint = f"http://{settings.SANDBOX_HOST}:9385"
@@ -187,6 +187,7 @@ class SelfManagedProvider(SandboxProvider):
                 )
 
             result = response.json()
+            structured_result = result.get("result") or {}
 
             return ExecutionResult(
                 stdout=result.get("stdout", ""),
@@ -199,6 +200,10 @@ class SelfManagedProvider(SandboxProvider):
                     "memory_used_kb": result.get("memory_used_kb"),
                     "detail": result.get("detail"),
                     "instance_id": instance_id,
+                    "artifacts": result.get("artifacts", []),
+                    "result_present": structured_result.get("present", False),
+                    "result_value": structured_result.get("value"),
+                    "result_type": structured_result.get("type"),
                 }
             )
 

agent/sandbox/pyproject.toml

@@ -8,7 +8,7 @@ dependencies = [
     "fastapi>=0.115.12",
     "httpx>=0.28.1",
     "pydantic>=2.11.4",
-    "requests>=2.32.3",
+    "requests>=2.32.4",
     "slowapi>=0.1.9",
     "uvicorn>=0.34.2",
 ]

agent/sandbox/sandbox_base_image/nodejs/package-lock.json

@@ -19,13 +19,13 @@
       "license": "MIT"
     },
     "node_modules/axios": {
-      "version": "1.12.0",
-      "resolved": "https://registry.npmjs.org/axios/-/axios-1.12.0.tgz",
-      "integrity": "sha512-oXTDccv8PcfjZmPGlWsPSwtOJCZ/b6W5jAMCNcfwJbCzDckwG0jrYJFaWH1yvivfCXjVzV/SPDEhMB3Q+DSurg==",
+      "version": "1.13.6",
+      "resolved": "https://registry.npmjs.org/axios/-/axios-1.13.6.tgz",
+      "integrity": "sha512-ChTCHMouEe2kn713WHbQGcuYrr6fXTBiu460OTwWrWob16g1bXn4vtz07Ope7ewMozJAnEquLk5lWQWtBig9DQ==",
       "license": "MIT",
       "dependencies": {
-        "follow-redirects": "^1.15.6",
-        "form-data": "^4.0.4",
+        "follow-redirects": "^1.15.11",
+        "form-data": "^4.0.5",
         "proxy-from-env": "^1.1.0"
       }
     },
@@ -123,9 +123,9 @@
       }
     },
     "node_modules/follow-redirects": {
-      "version": "1.15.9",
-      "resolved": "https://registry.npmmirror.com/follow-redirects/-/follow-redirects-1.15.9.tgz",
-      "integrity": "sha512-gew4GsXizNgdoRyqmyfMHyAmXsZDk6mHkSxZFCzW9gwlbtOW44CDtYavM+y+72qD/Vq2l550kMF52DT8fOLJqQ==",
+      "version": "1.15.11",
+      "resolved": "https://registry.npmjs.org/follow-redirects/-/follow-redirects-1.15.11.tgz",
+      "integrity": "sha512-deG2P0JfjrTxl50XGCDyfI97ZGVCxIpfKYmfyrQ54n5FO/0gfIES8C/Psl6kWVDolizcaaxZJnTS0QSMxvnsBQ==",
       "funding": [
         {
           "type": "individual",
@@ -143,9 +143,9 @@
       }
     },
     "node_modules/form-data": {
-      "version": "4.0.4",
-      "resolved": "https://registry.npmjs.org/form-data/-/form-data-4.0.4.tgz",
-      "integrity": "sha512-KrGhL9Q4zjj0kiUt5OO4Mr/A/jlI2jDYs5eHBpYHPcBEVSiipAvn2Ko2HnPe20rmcuuvMHNdZFp+4IlGTMF0Ow==",
+      "version": "4.0.5",
+      "resolved": "https://registry.npmjs.org/form-data/-/form-data-4.0.5.tgz",
+      "integrity": "sha512-8RipRLol37bNs2bhoV67fiTEvdTrbMUYcFTiy3+wuuOnUog2QBHCZWXDRijWQfAkhBj2Uf5UnVaiWwA5vdd82w==",
       "license": "MIT",
       "dependencies": {
         "asynckit": "^0.4.0",

agent/sandbox/sandbox_base_image/python/Dockerfile

@@ -2,12 +2,17 @@ FROM python:3.11-slim-bookworm
 
 COPY --from=ghcr.io/astral-sh/uv:0.7.5 /uv /uvx /bin/
 ENV UV_INDEX_URL=https://pypi.tuna.tsinghua.edu.cn/simple
+ENV MPLBACKEND=Agg
+ENV MPLCONFIGDIR=/tmp/matplotlib
+ENV MATPLOTLIBRC=/usr/local/etc/matplotlibrc
 
 COPY requirements.txt .
+COPY matplotlibrc /usr/local/etc/matplotlibrc
 
 RUN grep -rl 'deb.debian.org' /etc/apt/ | xargs sed -i 's|http[s]*://deb.debian.org|https://mirrors.tuna.tsinghua.edu.cn|g' && \
     apt-get update && \
     apt-get install -y curl gcc && \
+    mkdir -p /tmp/matplotlib && \
     uv pip install --system -r requirements.txt
 
 WORKDIR /workspace

agent/sandbox/sandbox_base_image/python/matplotlibrc

@@ -0,0 +1,11 @@
+## RAGFlow sandbox – matplotlib defaults
+## Only overrides are listed; all other settings use matplotlib built-in defaults.
+
+# Prefer CJK-capable fonts so Chinese / Japanese / Korean text renders correctly.
+# matplotlib silently skips fonts that are not installed, falling back to the
+# next entry in the list, so this is safe even without any CJK font package.
+font.family: sans-serif
+font.sans-serif: Noto Sans CJK SC, Noto Sans CJK TC, Noto Sans CJK JP, Noto Sans CJK KR, Source Han Sans SC, Source Han Sans CN, WenQuanYi Zen Hei, Microsoft YaHei, SimHei, PingFang SC, Heiti SC, STHeiti, Arial Unicode MS, DejaVu Sans, Bitstream Vera Sans, Computer Modern Sans Serif, Lucida Grande, Verdana, Geneva, Lucid, Arial, Helvetica, Avant Garde, sans-serif
+
+# Use ASCII hyphen-minus for the minus sign so it renders correctly with any font.
+axes.unicode_minus: False

agent/sandbox/sandbox_base_image/python/requirements.txt

@@ -1,3 +1,4 @@
 numpy
 pandas
+matplotlib
 requests

agent/sandbox/sandbox_spec.md

@@ -654,7 +654,7 @@ class AliyunCodeInterpreterProvider(SandboxProvider):
                 "type": "string",
                 "required": True,
                 "label": "Account ID",
-                "description": "Aliyun primary account ID (主账号ID), required for API calls"
+                "description": "Aliyun primary account ID, required for API calls"
             },
             "region": {
                 "type": "string",
@@ -1739,8 +1739,9 @@ def execute_code(
 
 1. **Self-managed provider** ([self_managed.py:164](agent/sandbox/providers/self_managed.py:164)):
    - Passes arguments via HTTP API: `"arguments": arguments or {}`
-   - executor_manager receives and passes to code via command line
-   - Runner script: `args = json.loads(sys.argv[1])` then `result = main(**args)`
+   - executor_manager writes `args.json` into the per-task workspace
+   - Runner script loads arguments from `args.json`
+   - Python runner calls `main(**args)` and JavaScript runner calls `main(args)`
 
 2. **Aliyun Code Interpreter** ([aliyun_codeinterpreter.py:260-275](agent/sandbox/providers/aliyun_codeinterpreter.py:260-275)):
    - Wraps user code to call `main(**arguments)` or `main()` if no arguments

agent/sandbox/tests/MIGRATION_GUIDE.md

@@ -1,53 +1,53 @@
-# Aliyun Code Interpreter Provider - 使用官方 SDK
+# Aliyun Code Interpreter Provider - Using the Official SDK
 
-## 重要变更
+## Important Changes
 
-### 官方资源
+### Official Resources
 - **Code Interpreter API**: https://help.aliyun.com/zh/functioncompute/fc/sandbox-sandbox-code-interepreter
-- **官方 SDK**: https://github.com/Serverless-Devs/agentrun-sdk-python
-- **SDK 文档**: https://docs.agent.run
+- **Official SDK**: https://github.com/Serverless-Devs/agentrun-sdk-python
+- **SDK Documentation**: https://docs.agent.run
 
-## 使用官方 SDK 的优势
+## Advantages of Using the Official SDK
 
-从手动 HTTP 请求迁移到官方 SDK (`agentrun-sdk`) 有以下优势：
+Migrating from manual HTTP requests to the official SDK (`agentrun-sdk`) offers the following benefits:
 
-### 1. **自动签名认证**
-- SDK 自动处理 Aliyun API 签名（无需手动实现 `Authorization` 头）
-- 支持多种认证方式：AccessKey、STS Token
-- 自动读取环境变量
+### 1. **Automatic Signature Authentication**
+- The SDK automatically handles Aliyun API signing (no need to manually implement `Authorization` headers)
+- Supports multiple authentication methods: AccessKey, STS Token
+- Automatically reads environment variables
 
-### 2. **简化的 API**
+### 2. **Simplified API**
 ```python
-# 旧实现（手动 HTTP 请求）
+# Old implementation (manual HTTP requests)
 response = requests.post(
     f"{DATA_ENDPOINT}/sandboxes/{sandbox_id}/execute",
     headers={"X-Acs-Parent-Id": account_id},
     json={"code": code, "language": "python"}
 )
 
-# 新实现（使用 SDK）
+# New implementation (using SDK)
 sandbox = CodeInterpreterSandbox(template_name="python-sandbox", config=config)
 result = sandbox.context.execute(code="print('hello')")
 ```
 
-### 3. **更好的错误处理**
-- 结构化的异常类型 (`ServerError`)
-- 自动重试机制
-- 详细的错误信息
+### 3. **Better Error Handling**
+- Structured exception types (`ServerError`)
+- Automatic retry mechanism
+- Detailed error messages
 
-## 主要变更
+## Key Changes
 
-### 1. 文件重命名
+### 1. File Renames
 
-| 旧文件名 | 新文件名 | 说明 |
+| Old Filename | New Filename | Description |
 |---------|---------|------|
-| `aliyun_opensandbox.py` | `aliyun_codeinterpreter.py` | 提供商实现 |
-| `test_aliyun_provider.py` | `test_aliyun_codeinterpreter.py` | 单元测试 |
-| `test_aliyun_integration.py` | `test_aliyun_codeinterpreter_integration.py` | 集成测试 |
+| `aliyun_opensandbox.py` | `aliyun_codeinterpreter.py` | Provider implementation |
+| `test_aliyun_provider.py` | `test_aliyun_codeinterpreter.py` | Unit tests |
+| `test_aliyun_integration.py` | `test_aliyun_codeinterpreter_integration.py` | Integration tests |
 
-### 2. 配置字段变更
+### 2. Configuration Field Changes
 
-#### 旧配置（OpenSandbox）
+#### Old Configuration (OpenSandbox)
 ```json
 {
   "access_key_id": "LTAI5t...",
@@ -57,59 +57,59 @@ result = sandbox.context.execute(code="print('hello')")
 }
 ```
 
-#### 新配置（Code Interpreter）
+#### New Configuration (Code Interpreter)
 ```json
 {
   "access_key_id": "LTAI5t...",
   "access_key_secret": "...",
-  "account_id": "1234567890...",  // 新增：阿里云主账号ID（必需）
+  "account_id": "1234567890...",  // New: Aliyun primary account ID (required)
   "region": "cn-hangzhou",
-  "template_name": "python-sandbox",  // 新增：沙箱模板名称
-  "timeout": 30  // 最大 30 秒（硬限制）
+  "template_name": "python-sandbox",  // New: sandbox template name
+  "timeout": 30  // Max 30 seconds (hard limit)
 }
 ```
 
-### 3. 关键差异
+### 3. Key Differences
 
-| 特性 | OpenSandbox | Code Interpreter |
+| Feature | OpenSandbox | Code Interpreter |
 |------|-------------|-----------------|
-| **API 端点** | `opensandbox.{region}.aliyuncs.com` | `agentrun.{region}.aliyuncs.com` (控制面) |
-| **API 版本** | `2024-01-01` | `2025-09-10` |
-| **认证** | 需要 AccessKey | 需要 AccessKey + 主账号ID |
-| **请求头** | 标准签名 | 需要 `X-Acs-Parent-Id` 头 |
-| **超时限制** | 可配置 | **最大 30 秒**（硬限制） |
-| **上下文** | 不支持 | 支持上下文（Jupyter kernel） |
+| **API Endpoint** | `opensandbox.{region}.aliyuncs.com` | `agentrun.{region}.aliyuncs.com` (control plane) |
+| **API Version** | `2024-01-01` | `2025-09-10` |
+| **Authentication** | AccessKey required | AccessKey + primary account ID required |
+| **Request Headers** | Standard signature | Requires `X-Acs-Parent-Id` header |
+| **Timeout Limit** | Configurable | **Max 30 seconds** (hard limit) |
+| **Context** | Not supported | Supports context (Jupyter kernel) |
 
-### 4. API 调用方式变更
+### 4. API Call Changes
 
-#### 旧实现（假设的 OpenSandbox）
+#### Old Implementation (assumed OpenSandbox)
 ```python
-# 单一端点
+# Single endpoint
 API_ENDPOINT = "https://opensandbox.cn-hangzhou.aliyuncs.com"
 
-# 简单的请求/响应
+# Simple request/response
 response = requests.post(
     f"{API_ENDPOINT}/execute",
     json={"code": "print('hello')", "language": "python"}
 )
 ```
 
-#### 新实现（Code Interpreter）
+#### New Implementation (Code Interpreter)
 ```python
-# 控制面 API - 管理沙箱生命周期
+# Control plane API - manage sandbox lifecycle
 CONTROL_ENDPOINT = "https://agentrun.cn-hangzhou.aliyuncs.com/2025-09-10"
 
-# 数据面 API - 执行代码
+# Data plane API - execute code
 DATA_ENDPOINT = "https://{account_id}.agentrun-data.cn-hangzhou.aliyuncs.com"
 
-# 创建沙箱（控制面）
+# Create sandbox (control plane)
 response = requests.post(
     f"{CONTROL_ENDPOINT}/sandboxes",
     headers={"X-Acs-Parent-Id": account_id},
     json={"templateName": "python-sandbox"}
 )
 
-# 执行代码（数据面）
+# Execute code (data plane)
 response = requests.post(
     f"{DATA_ENDPOINT}/sandboxes/{sandbox_id}/execute",
     headers={"X-Acs-Parent-Id": account_id},
@@ -117,13 +117,13 @@ response = requests.post(
 )
 ```
 
-### 5. 迁移步骤
+### 5. Migration Steps
 
-#### 步骤 1: 更新配置
+#### Step 1: Update Configuration
 
-如果您之前使用的是 `aliyun_opensandbox`：
+If you were previously using `aliyun_opensandbox`:
 
-**旧配置**:
+**Old configuration**:
 ```json
 {
   "name": "sandbox.provider_type",
@@ -131,7 +131,7 @@ response = requests.post(
 }
 ```
 
-**新配置**:
+**New configuration**:
 ```json
 {
   "name": "sandbox.provider_type",
@@ -139,123 +139,123 @@ response = requests.post(
 }
 ```
 
-#### 步骤 2: 添加必需的 account_id
+#### Step 2: Add the Required account_id
 
-在 Aliyun 控制台右上角点击头像，获取主账号 ID：
-1. 登录 [阿里云控制台](https://ram.console.aliyun.com/manage/ak)
-2. 点击右上角头像
-3. 复制主账号 ID（16 位数字）
+Get your primary account ID from the Aliyun console:
+1. Log in to the [Aliyun Console](https://ram.console.aliyun.com/manage/ak)
+2. Click on your avatar in the top-right corner
+3. Copy the primary account ID (16-digit number)
 
-#### 步骤 3: 更新环境变量
+#### Step 3: Update Environment Variables
 
 ```bash
-# 新增必需的环境变量
+# New required environment variable
 export ALIYUN_ACCOUNT_ID="1234567890123456"
 
-# 其他环境变量保持不变
+# Other environment variables remain unchanged
 export ALIYUN_ACCESS_KEY_ID="LTAI5t..."
 export ALIYUN_ACCESS_KEY_SECRET="..."
 export ALIYUN_REGION="cn-hangzhou"
 ```
 
-#### 步骤 4: 运行测试
+#### Step 4: Run Tests
 
 ```bash
-# 单元测试（不需要真实凭据）
+# Unit tests (no real credentials required)
 pytest agent/sandbox/tests/test_aliyun_codeinterpreter.py -v
 
-# 集成测试（需要真实凭据）
+# Integration tests (real credentials required)
 pytest agent/sandbox/tests/test_aliyun_codeinterpreter_integration.py -v -m integration
 ```
 
-## 文件变更清单
+## File Change Checklist
 
-### ✅ 已完成
+### ✅ Completed
 
-- [x] 创建 `aliyun_codeinterpreter.py` - 新的提供商实现
-- [x] 更新 `sandbox_spec.md` - 规范文档
-- [x] 更新 `admin/services.py` - 服务管理器
-- [x] 更新 `providers/__init__.py` - 包导出
-- [x] 创建 `test_aliyun_codeinterpreter.py` - 单元测试
-- [x] 创建 `test_aliyun_codeinterpreter_integration.py` - 集成测试
+- [x] Created `aliyun_codeinterpreter.py` - new provider implementation
+- [x] Updated `sandbox_spec.md` - specification document
+- [x] Updated `admin/services.py` - service manager
+- [x] Updated `providers/__init__.py` - package exports
+- [x] Created `test_aliyun_codeinterpreter.py` - unit tests
+- [x] Created `test_aliyun_codeinterpreter_integration.py` - integration tests
 
-### 📝 可选清理
+### 📝 Optional Cleanup
 
-如果您想删除旧的 OpenSandbox 实现：
+If you want to remove the old OpenSandbox implementation:
 
 ```bash
-# 删除旧文件（可选）
+# Remove old files (optional)
 rm agent/sandbox/providers/aliyun_opensandbox.py
 rm agent/sandbox/tests/test_aliyun_provider.py
 rm agent/sandbox/tests/test_aliyun_integration.py
 ```
 
-**注意**: 保留旧文件不会影响新功能，只是代码冗余。
+**Note**: Keeping the old files does not affect the new functionality; it just results in redundant code.
 
-## API 参考
+## API Reference
 
-### 控制面 API（沙箱管理）
+### Control Plane API (Sandbox Management)
 
-| 端点 | 方法 | 说明 |
+| Endpoint | Method | Description |
 |------|------|------|
-| `/sandboxes` | POST | 创建沙箱实例 |
-| `/sandboxes/{id}/stop` | POST | 停止实例 |
-| `/sandboxes/{id}` | DELETE | 删除实例 |
-| `/templates` | GET | 列出模板 |
+| `/sandboxes` | POST | Create a sandbox instance |
+| `/sandboxes/{id}/stop` | POST | Stop an instance |
+| `/sandboxes/{id}` | DELETE | Delete an instance |
+| `/templates` | GET | List templates |
 
-### 数据面 API（代码执行）
+### Data Plane API (Code Execution)
 
-| 端点 | 方法 | 说明 |
+| Endpoint | Method | Description |
 |------|------|------|
-| `/sandboxes/{id}/execute` | POST | 执行代码（简化版） |
-| `/sandboxes/{id}/contexts` | POST | 创建上下文 |
-| `/sandboxes/{id}/contexts/{ctx_id}/execute` | POST | 在上下文中执行 |
-| `/sandboxes/{id}/health` | GET | 健康检查 |
-| `/sandboxes/{id}/files` | GET/POST | 文件读写 |
-| `/sandboxes/{id}/processes/cmd` | POST | 执行 Shell 命令 |
+| `/sandboxes/{id}/execute` | POST | Execute code (simplified) |
+| `/sandboxes/{id}/contexts` | POST | Create a context |
+| `/sandboxes/{id}/contexts/{ctx_id}/execute` | POST | Execute within a context |
+| `/sandboxes/{id}/health` | GET | Health check |
+| `/sandboxes/{id}/files` | GET/POST | File read/write |
+| `/sandboxes/{id}/processes/cmd` | POST | Execute shell command |
 
-## 常见问题
+## FAQ
 
-### Q: 为什么要添加 account_id？
+### Q: Why is account_id required?
 
-**A**: Code Interpreter API 需要在请求头中提供 `X-Acs-Parent-Id`（阿里云主账号ID）进行身份验证。这是 Aliyun Code Interpreter API 的必需参数。
+**A**: The Code Interpreter API requires the `X-Acs-Parent-Id` (Aliyun primary account ID) header for authentication. This is a required parameter for the Aliyun Code Interpreter API.
 
-### Q: 30 秒超时限制可以绕过吗？
+### Q: Can the 30-second timeout limit be bypassed?
 
-**A**: 不可以。这是 Aliyun Code Interpreter 的**硬限制**，无法通过配置或请求参数绕过。如果代码执行时间超过 30 秒，请考虑：
-1. 优化代码逻辑
-2. 分批处理数据
-3. 使用上下文保持状态
+**A**: No. This is a **hard limit** of Aliyun Code Interpreter and cannot be bypassed through configuration or request parameters. If your code execution exceeds 30 seconds, consider:
+1. Optimizing the code logic
+2. Processing data in batches
+3. Using contexts to maintain state
 
-### Q: 旧的 OpenSandbox 配置还能用吗？
+### Q: Can the old OpenSandbox configuration still be used?
 
-**A**: 不能。OpenSandbox 和 Code Interpreter 是两个不同的服务，API 不兼容。必须迁移到新的配置格式。
+**A**: No. OpenSandbox and Code Interpreter are two different services with incompatible APIs. You must migrate to the new configuration format.
 
-### Q: 如何获取阿里云主账号 ID？
+### Q: How do I get the Aliyun primary account ID?
 
 **A**:
-1. 登录阿里云控制台
-2. 点击右上角的头像
-3. 在弹出的信息中可以看到"主账号ID"
+1. Log in to the Aliyun console
+2. Click on your avatar in the top-right corner
+3. The primary account ID will be displayed in the popup
 
-### Q: 迁移后会影响现有功能吗？
+### Q: Will the migration affect existing functionality?
 
 **A**:
-- **自我管理提供商（self_managed）**: 不受影响
-- **E2B 提供商**: 不受影响
-- **Aliyun 提供商**: 需要更新配置并重新测试
+- **Self-managed provider (self_managed)**: Not affected
+- **E2B provider**: Not affected
+- **Aliyun provider**: Configuration update and re-testing required
 
-## 相关文档
+## Related Documentation
 
-- [官方文档](https://help.aliyun.com/zh/functioncompute/fc/sandbox-sandbox-code-interepreter)
-- [sandbox 规范](../docs/develop/sandbox_spec.md)
-- [测试指南](./README.md)
-- [快速开始](./QUICKSTART.md)
+- [Official Documentation](https://help.aliyun.com/zh/functioncompute/fc/sandbox-sandbox-code-interepreter)
+- [Sandbox Specification](../docs/develop/sandbox_spec.md)
+- [Testing Guide](./README.md)
+- [Quick Start](./QUICKSTART.md)
 
-## 技术支持
+## Support
 
-如有问题，请：
-1. 查看官方文档
-2. 检查配置是否正确
-3. 查看测试输出中的错误信息
-4. 联系 RAGFlow 团队
+If you have any issues:
+1. Review the official documentation
+2. Verify the configuration is correct
+3. Check the error messages in the test output
+4. Contact the RAGFlow team

agent/sandbox/tests/QUICKSTART.md

@@ -1,45 +1,45 @@
-# Aliyun OpenSandbox Provider - 快速测试指南
+# Aliyun OpenSandbox Provider - Quick Test Guide
 
-## 测试说明
+## Test Overview
 
-### 1. 单元测试（不需要真实凭据）
+### 1. Unit Tests (No Credentials Required)
 
-单元测试使用 mock，**不需要**真实的 Aliyun 凭据，可以随时运行。
+Unit tests use mocks and do **not** require real Aliyun credentials; they can be run at any time.
 
 ```bash
-# 运行 Aliyun 提供商的单元测试
+# Run unit tests for the Aliyun provider
 pytest agent/sandbox/tests/test_aliyun_provider.py -v
 
-# 预期输出：
+# Expected output:
 # test_aliyun_provider.py::TestAliyunOpenSandboxProvider::test_provider_initialization PASSED
 # test_aliyun_provider.py::TestAliyunOpenSandboxProvider::test_initialize_success PASSED
 # ...
 # ========================= 48 passed in 2.34s ==========================
 ```
 
-### 2. 集成测试（需要真实凭据）
+### 2. Integration Tests (Real Credentials Required)
 
-集成测试会调用真实的 Aliyun API，需要配置凭据。
+Integration tests call the real Aliyun API and require credentials to be configured.
 
-#### 步骤 1: 配置环境变量
+#### Step 1: Configure Environment Variables
 
 ```bash
-export ALIYUN_ACCESS_KEY_ID="LTAI5t..."  # 替换为真实的 Access Key ID
-export ALIYUN_ACCESS_KEY_SECRET="..."     # 替换为真实的 Access Key Secret
-export ALIYUN_REGION="cn-hangzhou"        # 可选，默认为 cn-hangzhou
+export ALIYUN_ACCESS_KEY_ID="LTAI5t..."  # Replace with your real Access Key ID
+export ALIYUN_ACCESS_KEY_SECRET="..."     # Replace with your real Access Key Secret
+export ALIYUN_REGION="cn-hangzhou"        # Optional, defaults to cn-hangzhou
 ```
 
-#### 步骤 2: 运行集成测试
+#### Step 2: Run Integration Tests
 
 ```bash
-# 运行所有集成测试
+# Run all integration tests
 pytest agent/sandbox/tests/test_aliyun_integration.py -v -m integration
 
-# 运行特定测试
+# Run a specific test
 pytest agent/sandbox/tests/test_aliyun_integration.py::TestAliyunOpenSandboxIntegration::test_health_check -v
 ```
 
-#### 步骤 3: 预期输出
+#### Step 3: Expected Output
 
 ```
 test_aliyun_integration.py::TestAliyunOpenSandboxIntegration::test_initialize_provider PASSED
@@ -49,130 +49,130 @@ test_aliyun_integration.py::TestAliyunOpenSandboxIntegration::test_execute_pytho
 ========================== 10 passed in 15.67s ==========================
 ```
 
-### 3. 测试场景
+### 3. Test Scenarios
 
-#### 基础功能测试
+#### Basic Functionality Tests
 
 ```bash
-# 健康检查
+# Health check
 pytest agent/sandbox/tests/test_aliyun_integration.py::TestAliyunOpenSandboxIntegration::test_health_check -v
 
-# 创建实例
+# Create instance
 pytest agent/sandbox/tests/test_aliyun_integration.py::TestAliyunOpenSandboxIntegration::test_create_python_instance -v
 
-# 执行代码
+# Execute code
 pytest agent/sandbox/tests/test_aliyun_integration.py::TestAliyunOpenSandboxIntegration::test_execute_python_code -v
 
-# 销毁实例
+# Destroy instance
 pytest agent/sandbox/tests/test_aliyun_integration.py::TestAliyunOpenSandboxIntegration::test_destroy_instance -v
 ```
 
-#### 错误处理测试
+#### Error Handling Tests
 
 ```bash
-# 代码执行错误
+# Code execution error
 pytest agent/sandbox/tests/test_aliyun_integration.py::TestAliyunOpenSandboxIntegration::test_execute_python_code_with_error -v
 
-# 超时处理
+# Timeout handling
 pytest agent/sandbox/tests/test_aliyun_integration.py::TestAliyunOpenSandboxIntegration::test_execute_python_code_timeout -v
 ```
 
-#### 真实场景测试
+#### Real-World Scenario Tests
 
 ```bash
-# 数据处理工作流
+# Data processing workflow
 pytest agent/sandbox/tests/test_aliyun_integration.py::TestAliyunRealWorldScenarios::test_data_processing_workflow -v
 
-# 字符串操作
+# String manipulation
 pytest agent/sandbox/tests/test_aliyun_integration.py::TestAliyunRealWorldScenarios::test_string_manipulation -v
 
-# 多次执行
+# Multiple executions
 pytest agent/sandbox/tests/test_aliyun_integration.py::TestAliyunRealWorldScenarios::test_multiple_executions_same_instance -v
 ```
 
-## 常见问题
+## FAQ
 
-### Q: 没有凭据怎么办？
+### Q: What if I don't have credentials?
 
-**A:** 运行单元测试即可，不需要真实凭据：
+**A:** Just run the unit tests — no real credentials needed:
 ```bash
 pytest agent/sandbox/tests/test_aliyun_provider.py -v
 ```
 
-### Q: 如何跳过集成测试？
+### Q: How do I skip integration tests?
 
-**A:** 使用 pytest 标记跳过：
+**A:** Use pytest markers to skip them:
 ```bash
-# 只运行单元测试，跳过集成测试
+# Run only unit tests, skip integration tests
 pytest agent/sandbox/tests/ -v -m "not integration"
 ```
 
-### Q: 集成测试失败怎么办？
+### Q: What should I do if integration tests fail?
 
-**A:** 检查以下几点：
+**A:** Check the following:
 
-1. **凭据是否正确**
+1. **Are the credentials correct?**
    ```bash
    echo $ALIYUN_ACCESS_KEY_ID
    echo $ALIYUN_ACCESS_KEY_SECRET
    ```
 
-2. **网络连接是否正常**
+2. **Is the network connection working?**
    ```bash
    curl -I https://opensandbox.cn-hangzhou.aliyuncs.com
    ```
 
-3. **是否有 OpenSandbox 服务权限**
-   - 登录阿里云控制台
-   - 检查是否已开通 OpenSandbox 服务
-   - 检查 AccessKey 权限
+3. **Do you have OpenSandbox service permissions?**
+   - Log in to the Aliyun console
+   - Check if the OpenSandbox service is enabled
+   - Verify AccessKey permissions
 
-4. **查看详细错误信息**
+4. **View detailed error messages:**
    ```bash
    pytest agent/sandbox/tests/test_aliyun_integration.py -v -s
    ```
 
-### Q: 测试超时怎么办？
+### Q: What should I do if tests time out?
 
-**A:** 增加超时时间或检查网络：
+**A:** Increase the timeout or check network connectivity:
 ```bash
-# 使用更长的超时
+# Use a longer timeout
 pytest agent/sandbox/tests/test_aliyun_integration.py -v --timeout=60
 ```
 
-## 测试命令速查表
+## Quick Reference: Test Commands
 
-| 命令 | 说明 | 需要凭据 |
+| Command | Description | Credentials Required |
 |------|------|---------|
-| `pytest agent/sandbox/tests/test_aliyun_provider.py -v` | 单元测试 | ❌ |
-| `pytest agent/sandbox/tests/test_aliyun_integration.py -v` | 集成测试 | ✅ |
-| `pytest agent/sandbox/tests/ -v -m "not integration"` | 仅单元测试 | ❌ |
-| `pytest agent/sandbox/tests/ -v -m integration` | 仅集成测试 | ✅ |
-| `pytest agent/sandbox/tests/ -v` | 所有测试 | 部分需要 |
+| `pytest agent/sandbox/tests/test_aliyun_provider.py -v` | Unit tests | ❌ |
+| `pytest agent/sandbox/tests/test_aliyun_integration.py -v` | Integration tests | ✅ |
+| `pytest agent/sandbox/tests/ -v -m "not integration"` | Unit tests only | ❌ |
+| `pytest agent/sandbox/tests/ -v -m integration` | Integration tests only | ✅ |
+| `pytest agent/sandbox/tests/ -v` | All tests | Partially required |
 
-## 获取 Aliyun 凭据
+## Getting Aliyun Credentials
 
-1. 访问 [阿里云控制台](https://ram.console.aliyun.com/manage/ak)
-2. 创建 AccessKey
-3. 保存 AccessKey ID 和 AccessKey Secret
-4. 设置环境变量
+1. Visit the [Aliyun Console](https://ram.console.aliyun.com/manage/ak)
+2. Create an AccessKey
+3. Save your AccessKey ID and AccessKey Secret
+4. Set the environment variables
 
-⚠️ **安全提示：**
-- 不要在代码中硬编码凭据
-- 使用环境变量或配置文件
-- 定期轮换 AccessKey
-- 限制 AccessKey 权限
+⚠️ **Security Tips:**
+- Do not hardcode credentials in your code
+- Use environment variables or configuration files
+- Rotate AccessKeys regularly
+- Restrict AccessKey permissions
 
-## 下一步
+## Next Steps
 
-1. ✅ **运行单元测试** - 验证代码逻辑
-2. 🔧 **配置凭据** - 设置环境变量
-3. 🚀 **运行集成测试** - 测试真实 API
-4. 📊 **查看结果** - 确保所有测试通过
-5. 🎯 **集成到系统** - 使用 admin API 配置提供商
+1. ✅ **Run unit tests** - Verify code logic
+2. 🔧 **Configure credentials** - Set environment variables
+3. 🚀 **Run integration tests** - Test the real API
+4. 📊 **Review results** - Ensure all tests pass
+5. 🎯 **Integrate into your system** - Configure the provider via the admin API
 
-## 需要帮助？
+## Need Help?
 
-- 查看 [完整文档](README.md)
-- 检查 [sandbox 规范](../../../../../docs/develop/sandbox_spec.md)
-- 联系 RAGFlow 团队
+- See the [full documentation](README.md)
+- Check the [sandbox specification](../../../../../docs/develop/sandbox_spec.md)
+- Contact the RAGFlow team

agent/sandbox/tests/test_aliyun_codeinterpreter.py

@@ -101,13 +101,15 @@ class TestAliyunCodeInterpreterProvider:
         assert provider.region == "cn-hangzhou"
         assert provider.template_name == ""
 
-    @patch("agent.sandbox.providers.aliyun_codeinterpreter.CodeInterpreterSandbox")
-    def test_create_instance_python(self, mock_sandbox_class):
+    @patch("agent.sandbox.providers.aliyun_codeinterpreter.Template")
+    @patch("agent.sandbox.providers.aliyun_codeinterpreter.Sandbox")
+    def test_create_instance_python(self, mock_sandbox_class, mock_template):
         """Test creating a Python instance."""
         # Mock successful instance creation
         mock_sandbox = MagicMock()
         mock_sandbox.sandbox_id = "01JCED8Z9Y6XQVK8M2NRST5WXY"
-        mock_sandbox_class.return_value = mock_sandbox
+        mock_sandbox_class.create.return_value = mock_sandbox
+        mock_template.get_by_name.return_value = MagicMock()
 
         provider = AliyunCodeInterpreterProvider()
         provider._initialized = True
@@ -119,12 +121,14 @@ class TestAliyunCodeInterpreterProvider:
         assert instance.status == "READY"
         assert instance.metadata["language"] == "python"
 
-    @patch("agent.sandbox.providers.aliyun_codeinterpreter.CodeInterpreterSandbox")
-    def test_create_instance_javascript(self, mock_sandbox_class):
+    @patch("agent.sandbox.providers.aliyun_codeinterpreter.Template")
+    @patch("agent.sandbox.providers.aliyun_codeinterpreter.Sandbox")
+    def test_create_instance_javascript(self, mock_sandbox_class, mock_template):
         """Test creating a JavaScript instance."""
         mock_sandbox = MagicMock()
         mock_sandbox.sandbox_id = "01JCED8Z9Y6XQVK8M2NRST5WXY"
-        mock_sandbox_class.return_value = mock_sandbox
+        mock_sandbox_class.create.return_value = mock_sandbox
+        mock_template.get_by_name.return_value = MagicMock()
 
         provider = AliyunCodeInterpreterProvider()
         provider._initialized = True
@@ -141,7 +145,7 @@ class TestAliyunCodeInterpreterProvider:
         with pytest.raises(RuntimeError, match="Provider not initialized"):
             provider.create_instance("python")
 
-    @patch("agent.sandbox.providers.aliyun_codeinterpreter.CodeInterpreterSandbox")
+    @patch("agent.sandbox.providers.aliyun_codeinterpreter.Sandbox")
     def test_execute_code_success(self, mock_sandbox_class):
         """Test successful code execution."""
         # Mock sandbox instance
@@ -150,7 +154,7 @@ class TestAliyunCodeInterpreterProvider:
             "results": [{"type": "stdout", "text": "Hello, World!"}, {"type": "result", "text": "None"}, {"type": "endOfExecution", "status": "ok"}],
             "contextId": "kernel-12345-67890",
         }
-        mock_sandbox_class.return_value = mock_sandbox
+        mock_sandbox_class.connect.return_value = mock_sandbox
 
         provider = AliyunCodeInterpreterProvider()
         provider._initialized = True
@@ -163,14 +167,14 @@ class TestAliyunCodeInterpreterProvider:
         assert result.exit_code == 0
         assert result.execution_time > 0
 
-    @patch("agent.sandbox.providers.aliyun_codeinterpreter.CodeInterpreterSandbox")
+    @patch("agent.sandbox.providers.aliyun_codeinterpreter.Sandbox")
     def test_execute_code_timeout(self, mock_sandbox_class):
         """Test code execution timeout."""
         from agentrun.utils.exception import ServerError
 
         mock_sandbox = MagicMock()
         mock_sandbox.context.execute.side_effect = ServerError(408, "Request timeout")
-        mock_sandbox_class.return_value = mock_sandbox
+        mock_sandbox_class.connect.return_value = mock_sandbox
 
         provider = AliyunCodeInterpreterProvider()
         provider._initialized = True
@@ -179,14 +183,14 @@ class TestAliyunCodeInterpreterProvider:
         with pytest.raises(TimeoutError, match="Execution timed out"):
             provider.execute_code(instance_id="01JCED8Z9Y6XQVK8M2NRST5WXY", code="while True: pass", language="python", timeout=5)
 
-    @patch("agent.sandbox.providers.aliyun_codeinterpreter.CodeInterpreterSandbox")
+    @patch("agent.sandbox.providers.aliyun_codeinterpreter.Sandbox")
     def test_execute_code_with_error(self, mock_sandbox_class):
         """Test code execution with error."""
         mock_sandbox = MagicMock()
         mock_sandbox.context.execute.return_value = {
             "results": [{"type": "stderr", "text": "Traceback..."}, {"type": "error", "text": "NameError: name 'x' is not defined"}, {"type": "endOfExecution", "status": "error"}]
         }
-        mock_sandbox_class.return_value = mock_sandbox
+        mock_sandbox_class.connect.return_value = mock_sandbox
 
         provider = AliyunCodeInterpreterProvider()
         provider._initialized = True
@@ -197,6 +201,34 @@ class TestAliyunCodeInterpreterProvider:
         assert result.exit_code != 0
         assert len(result.stderr) > 0
 
+    @patch("agent.sandbox.providers.aliyun_codeinterpreter.Sandbox")
+    def test_execute_code_uses_structured_result_marker_for_async_javascript(self, mock_sandbox_class):
+        """Test JavaScript wrapper uses the structured result marker and awaits async main."""
+        mock_sandbox = MagicMock()
+        mock_sandbox.context.execute.return_value = {
+            "results": [{"type": "stdout", "text": "__RAGFLOW_RESULT__:eyJwcmVzZW50Ijp0cnVlLCJ2YWx1ZSI6eyJhIjoiYiJ9LCJ0eXBlIjoianNvbiJ9"}],
+            "contextId": "kernel-12345-67890",
+        }
+        mock_sandbox_class.connect.return_value = mock_sandbox
+
+        provider = AliyunCodeInterpreterProvider()
+        provider._initialized = True
+        provider._config = MagicMock()
+
+        result = provider.execute_code(
+            instance_id="01JCED8Z9Y6XQVK8M2NRST5WXY",
+            code="async function main(args) { return { a: 'b' }; }",
+            language="javascript",
+            timeout=10,
+        )
+
+        wrapped_code = mock_sandbox.context.execute.call_args.kwargs["code"]
+        assert "__RAGFLOW_RESULT__:" in wrapped_code
+        assert "await Promise.resolve(main(" in wrapped_code
+        assert result.metadata["result_present"] is True
+        assert result.metadata["result_value"] == {"a": "b"}
+        assert result.metadata["result_type"] == "json"
+
     def test_get_supported_languages(self):
         """Test getting supported languages."""
         provider = AliyunCodeInterpreterProvider()

agent/sandbox/tests/test_aliyun_codeinterpreter_integration.py

@@ -22,7 +22,7 @@ To run these tests, set the following environment variables:
 
     export AGENTRUN_ACCESS_KEY_ID="LTAI5t..."
     export AGENTRUN_ACCESS_KEY_SECRET="..."
-    export AGENTRUN_ACCOUNT_ID="1234567890..."  # Aliyun primary account ID (主账号ID)
+    export AGENTRUN_ACCOUNT_ID="1234567890..."  # Aliyun primary account ID
     export AGENTRUN_REGION="cn-hangzhou"  # Note: AGENTRUN_REGION (SDK will read this)
 
 Then run:

agent/sandbox/tests/test_providers.py

@@ -254,6 +254,41 @@ class TestSelfManagedProvider:
         assert result.metadata["status"] == "success"
         assert result.metadata["instance_id"] == "test-123"
 
+    @patch('requests.post')
+    def test_execute_code_maps_structured_result_into_metadata(self, mock_post):
+        """Test successful code execution with structured result envelope."""
+        mock_response = Mock()
+        mock_response.status_code = 200
+        mock_response.json.return_value = {
+            "status": "success",
+            "stdout": "debug line\n",
+            "stderr": "",
+            "exit_code": 0,
+            "time_used_ms": 100.0,
+            "memory_used_kb": 1024.0,
+            "result": {
+                "present": True,
+                "value": {"items": ["a", "b"]},
+                "type": "json",
+            },
+        }
+        mock_post.return_value = mock_response
+
+        provider = SelfManagedProvider()
+        provider._initialized = True
+
+        result = provider.execute_code(
+            instance_id="test-123",
+            code="def main(): return {'items': ['a', 'b']}",
+            language="python",
+            timeout=10
+        )
+
+        assert result.stdout == "debug line\n"
+        assert result.metadata["result_present"] is True
+        assert result.metadata["result_value"] == {"items": ["a", "b"]}
+        assert result.metadata["result_type"] == "json"
+
     @patch('requests.post')
     def test_execute_code_timeout(self, mock_post):
         """Test code execution timeout."""

agent/sandbox/tests/test_security.py

@@ -0,0 +1,55 @@
+#
+#  Copyright 2026 The InfiniFlow Authors. All Rights Reserved.
+#
+#  Licensed under the Apache License, Version 2.0 (the "License");
+#  you may not use this file except in compliance with the License.
+#  You may obtain a copy of the License at
+#
+#      http://www.apache.org/licenses/LICENSE-2.0
+#
+#  Unless required by applicable law or agreed to in writing, software
+#  distributed under the License is distributed on an "AS IS" BASIS,
+#  WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+#  See the License for the specific language governing permissions and
+#  limitations under the License.
+#
+import sys
+from pathlib import Path
+
+
+EXECUTOR_MANAGER_ROOT = Path(__file__).resolve().parents[1] / "executor_manager"
+if str(EXECUTOR_MANAGER_ROOT) not in sys.path:
+    sys.path.insert(0, str(EXECUTOR_MANAGER_ROOT))
+
+from models.enums import SupportLanguage  # noqa: E402
+from services.security import analyze_code_security  # noqa: E402
+
+
+def test_javascript_child_process_is_rejected():
+    is_safe, issues = analyze_code_security(
+        "const cp = require('child_process'); async function main() { return 'ok'; }",
+        SupportLanguage.NODEJS,
+    )
+
+    assert is_safe is False
+    assert any("child_process" in issue for issue, _ in issues)
+
+
+def test_javascript_eval_is_rejected():
+    is_safe, issues = analyze_code_security(
+        "async function main() { return eval('1+1'); }",
+        SupportLanguage.NODEJS,
+    )
+
+    assert is_safe is False
+    assert any("eval" in issue.lower() for issue, _ in issues)
+
+
+def test_javascript_safe_code_still_passes():
+    is_safe, issues = analyze_code_security(
+        "async function main(args) { return { answer: args.value ?? null }; }",
+        SupportLanguage.NODEJS,
+    )
+
+    assert is_safe is True
+    assert issues == []

agent/sandbox/uv.lock

@@ -1,6 +1,6 @@
 version = 1
 revision = 3
-requires-python = ">=3.10"
+requires-python = ">=3.12, <3.15"
 
 [[package]]
 name = "annotated-doc"
@@ -161,7 +161,7 @@ requires-dist = [
     { name = "fastapi", specifier = ">=0.115.12" },
     { name = "httpx", specifier = ">=0.28.1" },
     { name = "pydantic", specifier = ">=2.11.4" },
-    { name = "requests", specifier = ">=2.32.3" },
+    { name = "requests", specifier = ">=2.32.4" },
     { name = "slowapi", specifier = ">=0.1.9" },
     { name = "uvicorn", specifier = ">=0.34.2" },
 ]
@@ -313,7 +313,7 @@ wheels = [
 
 [[package]]
 name = "requests"
-version = "2.32.3"
+version = "2.32.5"
 source = { registry = "https://pypi.tuna.tsinghua.edu.cn/simple" }
 dependencies = [
     { name = "certifi" },
@@ -321,9 +321,9 @@ dependencies = [
     { name = "idna" },
     { name = "urllib3" },
 ]
-sdist = { url = "https://pypi.tuna.tsinghua.edu.cn/packages/63/70/2bf7780ad2d390a8d301ad0b550f1581eadbd9a20f896afe06353c2a2913/requests-2.32.3.tar.gz", hash = "sha256:55365417734eb18255590a9ff9eb97e9e1da868d4ccd6402399eaf68af20a760", size = 131218, upload-time = "2024-05-29T15:37:49.536Z" }
+sdist = { url = "https://pypi.tuna.tsinghua.edu.cn/packages/c9/74/b3ff8e6c8446842c3f5c837e9c3dfcfe2018ea6ecef224c710c85ef728f4/requests-2.32.5.tar.gz", hash = "sha256:dbba0bac56e100853db0ea71b82b4dfd5fe2bf6d3754a8893c3af500cec7d7cf", size = 134517, upload-time = "2025-08-18T20:46:02.573Z" }
 wheels = [
-    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/f9/9b/335f9764261e915ed497fcdeb11df5dfd6f7bf257d4a6a2a686d80da4d54/requests-2.32.3-py3-none-any.whl", hash = "sha256:70761cfe03c773ceb22aa2f671b4757976145175cdfca038c02654d061d6dcc6", size = 64928, upload-time = "2024-05-29T15:37:47.027Z" },
+    { url = "https://pypi.tuna.tsinghua.edu.cn/packages/1e/db/4254e3eabe8020b458f1a747140d32277ec7a271daf1d235b70dc0b4e6e3/requests-2.32.5-py3-none-any.whl", hash = "sha256:2462f94637a34fd532264295e186976db0f5d453d1cdd31473c85a6a161affb6", size = 64738, upload-time = "2025-08-18T20:46:00.542Z" },
 ]
 
 [[package]]

agent/templates/cv_analysis_and_candidate_evaluation.json

@@ -94,7 +94,7 @@
                                     "type": "integer"
                                 },
                                 "item": {
-                                    "type": "unkown"
+                                    "type": "unknown"
                                 }
                             }
                         }
@@ -252,7 +252,7 @@
                                         "type": "integer"
                                     },
                                     "item": {
-                                        "type": "unkown"
+                                        "type": "unknown"
                                     }
                                 }
                             },

agent/templates/knowledge_base_report_r.json

@@ -1,333 +0,0 @@
-{
-    "id": 21,
-    "title": {
-        "en": "Report Agent Using Knowledge Base",
-        "de": "Berichtsagent mit Wissensdatenbank",
-        "zh": "知识库检索智能体"},
-    "description": {
-        "en": "A report generation assistant using local knowledge base, with advanced capabilities in task planning, reasoning, and reflective analysis. Recommended for academic research paper Q&A",
-        "de": "Ein Berichtsgenerierungsassistent, der eine lokale Wissensdatenbank nutzt, mit erweiterten Fähigkeiten in Aufgabenplanung, Schlussfolgerung und reflektierender Analyse. Empfohlen für akademische Forschungspapier-Fragen und -Antworten.",
-        "zh": "一个使用本地知识库的报告生成助手，具备高级能力，包括任务规划、推理和反思性分析。推荐用于学术研究论文问答。"},
-    "canvas_type": "Recommended",
-    "dsl": {
-        "components": {
-            "Agent:NewPumasLick": {
-                "downstream": [
-                    "Message:OrangeYearsShine"
-                ],
-                "obj": {
-                    "component_name": "Agent",
-                    "params": {
-                        "delay_after_error": 1,
-                        "description": "",
-                        "exception_comment": "",
-                        "exception_default_value": "",
-                        "exception_goto": [],
-                        "exception_method": null,
-                        "frequencyPenaltyEnabled": false,
-                        "frequency_penalty": 0.5,
-                        "llm_id": "qwen3-235b-a22b-instruct-2507@Tongyi-Qianwen",
-                        "maxTokensEnabled": true,
-                        "max_retries": 3,
-                        "max_rounds": 3,
-                        "max_tokens": 128000,
-                        "mcp": [],
-                        "message_history_window_size": 12,
-                        "outputs": {
-                            "content": {
-                                "type": "string",
-                                "value": ""
-                            }
-                        },
-                        "parameter": "Precise",
-                        "presencePenaltyEnabled": false,
-                        "presence_penalty": 0.5,
-                        "prompts": [
-                            {
-                                "content": "# User Query\n {sys.query}",
-                                "role": "user"
-                            }
-                        ],
-                        "sys_prompt": "## Role & Task\nYou are a **\u201cKnowledge Base Retrieval Q\\&A Agent\u201d** whose goal is to break down the user\u2019s question into retrievable subtasks, and then produce a multi-source-verified, structured, and actionable research report using the internal knowledge base.\n## Execution Framework (Detailed Steps & Key Points)\n1. **Assessment & Decomposition**\n   * Actions:\n     * Automatically extract: main topic, subtopics, entities (people/organizations/products/technologies), time window, geographic/business scope.\n     * Output as a list: N facts/data points that must be collected (*N* ranges from 5\u201320 depending on question complexity).\n2. **Query Type Determination (Rule-Based)**\n   * Example rules:\n     * If the question involves a single issue but requests \u201cmethod comparison/multiple explanations\u201d \u2192 use **depth-first**.\n     * If the question can naturally be split into \u22653 independent sub-questions \u2192 use **breadth-first**.\n     * If the question can be answered by a single fact/specification/definition \u2192 use **simple query**.\n3. **Research Plan Formulation**\n   * Depth-first: define 3\u20135 perspectives (methodology/stakeholders/time dimension/technical route, etc.), assign search keywords, target document types, and output format for each perspective.\n   * Breadth-first: list subtasks, prioritize them, and assign search terms.\n   * Simple query: directly provide the search sentence and required fields.\n4. **Retrieval Execution**\n   * After retrieval: perform coverage check (does it contain the key facts?) and quality check (source diversity, authority, latest update time).\n   * If standards are not met, automatically loop: rewrite queries (synonyms/cross-domain terms) and retry \u22643 times, or flag as requiring external search.\n5. **Integration & Reasoning**\n   * Build the answer using a **fact\u2013evidence\u2013reasoning** chain. For each conclusion, attach 1\u20132 strongest pieces of evidence.\n---\n## Quality Gate Checklist (Verify at Each Stage)\n* **Stage 1 (Decomposition)**:\n  * [ ] Key concepts and expected outputs identified\n  * [ ] Required facts/data points listed\n* **Stage 2 (Retrieval)**:\n  * [ ] Meets quality standards (see above)\n  * [ ] If not met: execute query iteration\n* **Stage 3 (Generation)**:\n  * [ ] Each conclusion has at least one direct evidence source\n  * [ ] State assumptions/uncertainties\n  * [ ] Provide next-step suggestions or experiment/retrieval plans\n  * [ ] Final length and depth match user expectations (comply with word count/format if specified)\n---\n## Core Principles\n1. **Strict reliance on the knowledge base**: answers must be **fully bounded** by the content retrieved from the knowledge base.\n2. **No fabrication**: do not generate, infer, or create information that is not explicitly present in the knowledge base.\n3. **Accuracy first**: prefer incompleteness over inaccurate content.\n4. **Output format**:\n   * Hierarchically clear modular structure\n   * Logical grouping according to the MECE principle\n   * Professionally presented formatting\n   * Step-by-step cognitive guidance\n   * Reasonable use of headings and dividers for clarity\n   * *Italicize* key parameters\n   * **Bold** critical information\n5. **LaTeX formula requirements**:\n   * Inline formulas: start and end with `$`\n   * Block formulas: start and end with `$$`, each `$$` on its own line\n   * Block formula content must comply with LaTeX math syntax\n   * Verify formula correctness\n---\n## Additional Notes (Interaction & Failure Strategy)\n* If the knowledge base does not cover critical facts: explicitly inform the user (with sample wording)\n* For time-sensitive issues: enforce time filtering in the search request, and indicate the latest retrieval date in the answer.\n* Language requirement: answer in the user\u2019s preferred language\n",
-                        "temperature": "0.1",
-                        "temperatureEnabled": true,
-                        "tools": [
-                            {
-                                "component_name": "Retrieval",
-                                "name": "Retrieval",
-                                "params": {
-                                    "cross_languages": [],
-                                    "description": "",
-                                    "empty_response": "",
-                                    "kb_ids": [],
-                                    "keywords_similarity_weight": 0.7,
-                                    "outputs": {
-                                        "formalized_content": {
-                                            "type": "string",
-                                            "value": ""
-                                        }
-                                    },
-                                    "rerank_id": "",
-                                    "similarity_threshold": 0.2,
-                                    "top_k": 1024,
-                                    "top_n": 8,
-                                    "use_kg": false
-                                }
-                            }
-                        ],
-                        "topPEnabled": false,
-                        "top_p": 0.75,
-                        "user_prompt": "",
-                        "visual_files_var": ""
-                    }
-                },
-                "upstream": [
-                    "begin"
-                ]
-            },
-            "Message:OrangeYearsShine": {
-                "downstream": [],
-                "obj": {
-                    "component_name": "Message",
-                    "params": {
-                        "content": [
-                            "{Agent:NewPumasLick@content}"
-                        ]
-                    }
-                },
-                "upstream": [
-                    "Agent:NewPumasLick"
-                ]
-            },
-            "begin": {
-                "downstream": [
-                    "Agent:NewPumasLick"
-                ],
-                "obj": {
-                    "component_name": "Begin",
-                    "params": {
-                        "enablePrologue": true,
-                        "inputs": {},
-                        "mode": "conversational",
-                        "prologue": "\u4f60\u597d\uff01 \u6211\u662f\u4f60\u7684\u52a9\u7406\uff0c\u6709\u4ec0\u4e48\u53ef\u4ee5\u5e2e\u5230\u4f60\u7684\u5417\uff1f"
-                    }
-                },
-                "upstream": []
-            }
-        },
-        "globals": {
-            "sys.conversation_turns": 0,
-            "sys.files": [],
-            "sys.query": "",
-            "sys.user_id": ""
-        },
-        "graph": {
-            "edges": [
-                {
-                    "data": {
-                        "isHovered": false
-                    },
-                    "id": "xy-edge__beginstart-Agent:NewPumasLickend",
-                    "source": "begin",
-                    "sourceHandle": "start",
-                    "target": "Agent:NewPumasLick",
-                    "targetHandle": "end"
-                },
-                {
-                    "data": {
-                        "isHovered": false
-                    },
-                    "id": "xy-edge__Agent:NewPumasLickstart-Message:OrangeYearsShineend",
-                    "markerEnd": "logo",
-                    "source": "Agent:NewPumasLick",
-                    "sourceHandle": "start",
-                    "style": {
-                        "stroke": "rgba(91, 93, 106, 1)",
-                        "strokeWidth": 1
-                    },
-                    "target": "Message:OrangeYearsShine",
-                    "targetHandle": "end",
-                    "type": "buttonEdge",
-                    "zIndex": 1001
-                },
-                {
-                    "data": {
-                        "isHovered": false
-                    },
-                    "id": "xy-edge__Agent:NewPumasLicktool-Tool:AllBirdsNailend",
-                    "selected": false,
-                    "source": "Agent:NewPumasLick",
-                    "sourceHandle": "tool",
-                    "target": "Tool:AllBirdsNail",
-                    "targetHandle": "end"
-                }
-            ],
-            "nodes": [
-                {
-                    "data": {
-                        "form": {
-                            "enablePrologue": true,
-                            "inputs": {},
-                            "mode": "conversational",
-                            "prologue": "\u4f60\u597d\uff01 \u6211\u662f\u4f60\u7684\u52a9\u7406\uff0c\u6709\u4ec0\u4e48\u53ef\u4ee5\u5e2e\u5230\u4f60\u7684\u5417\uff1f"
-                        },
-                        "label": "Begin",
-                        "name": "begin"
-                    },
-                    "dragging": false,
-                    "id": "begin",
-                    "measured": {
-                        "height": 48,
-                        "width": 200
-                    },
-                    "position": {
-                        "x": -9.569875358221438,
-                        "y": 205.84018385864917
-                    },
-                    "selected": false,
-                    "sourcePosition": "left",
-                    "targetPosition": "right",
-                    "type": "beginNode"
-                },
-                {
-                    "data": {
-                        "form": {
-                            "content": [
-                                "{Agent:NewPumasLick@content}"
-                            ]
-                        },
-                        "label": "Message",
-                        "name": "Response"
-                    },
-                    "dragging": false,
-                    "id": "Message:OrangeYearsShine",
-                    "measured": {
-                        "height": 56,
-                        "width": 200
-                    },
-                    "position": {
-                        "x": 734.4061285881053,
-                        "y": 199.9706031723009
-                    },
-                    "selected": false,
-                    "sourcePosition": "right",
-                    "targetPosition": "left",
-                    "type": "messageNode"
-                },
-                {
-                    "data": {
-                        "form": {
-                            "delay_after_error": 1,
-                            "description": "",
-                            "exception_comment": "",
-                            "exception_default_value": "",
-                            "exception_goto": [],
-                            "exception_method": null,
-                            "frequencyPenaltyEnabled": false,
-                            "frequency_penalty": 0.5,
-                            "llm_id": "qwen3-235b-a22b-instruct-2507@Tongyi-Qianwen",
-                            "maxTokensEnabled": true,
-                            "max_retries": 3,
-                            "max_rounds": 3,
-                            "max_tokens": 128000,
-                            "mcp": [],
-                            "message_history_window_size": 12,
-                            "outputs": {
-                                "content": {
-                                    "type": "string",
-                                    "value": ""
-                                }
-                            },
-                            "parameter": "Precise",
-                            "presencePenaltyEnabled": false,
-                            "presence_penalty": 0.5,
-                            "prompts": [
-                                {
-                                    "content": "# User Query\n {sys.query}",
-                                    "role": "user"
-                                }
-                            ],
-                            "sys_prompt": "## Role & Task\nYou are a **\u201cKnowledge Base Retrieval Q\\&A Agent\u201d** whose goal is to break down the user\u2019s question into retrievable subtasks, and then produce a multi-source-verified, structured, and actionable research report using the internal knowledge base.\n## Execution Framework (Detailed Steps & Key Points)\n1. **Assessment & Decomposition**\n   * Actions:\n     * Automatically extract: main topic, subtopics, entities (people/organizations/products/technologies), time window, geographic/business scope.\n     * Output as a list: N facts/data points that must be collected (*N* ranges from 5\u201320 depending on question complexity).\n2. **Query Type Determination (Rule-Based)**\n   * Example rules:\n     * If the question involves a single issue but requests \u201cmethod comparison/multiple explanations\u201d \u2192 use **depth-first**.\n     * If the question can naturally be split into \u22653 independent sub-questions \u2192 use **breadth-first**.\n     * If the question can be answered by a single fact/specification/definition \u2192 use **simple query**.\n3. **Research Plan Formulation**\n   * Depth-first: define 3\u20135 perspectives (methodology/stakeholders/time dimension/technical route, etc.), assign search keywords, target document types, and output format for each perspective.\n   * Breadth-first: list subtasks, prioritize them, and assign search terms.\n   * Simple query: directly provide the search sentence and required fields.\n4. **Retrieval Execution**\n   * After retrieval: perform coverage check (does it contain the key facts?) and quality check (source diversity, authority, latest update time).\n   * If standards are not met, automatically loop: rewrite queries (synonyms/cross-domain terms) and retry \u22643 times, or flag as requiring external search.\n5. **Integration & Reasoning**\n   * Build the answer using a **fact\u2013evidence\u2013reasoning** chain. For each conclusion, attach 1\u20132 strongest pieces of evidence.\n---\n## Quality Gate Checklist (Verify at Each Stage)\n* **Stage 1 (Decomposition)**:\n  * [ ] Key concepts and expected outputs identified\n  * [ ] Required facts/data points listed\n* **Stage 2 (Retrieval)**:\n  * [ ] Meets quality standards (see above)\n  * [ ] If not met: execute query iteration\n* **Stage 3 (Generation)**:\n  * [ ] Each conclusion has at least one direct evidence source\n  * [ ] State assumptions/uncertainties\n  * [ ] Provide next-step suggestions or experiment/retrieval plans\n  * [ ] Final length and depth match user expectations (comply with word count/format if specified)\n---\n## Core Principles\n1. **Strict reliance on the knowledge base**: answers must be **fully bounded** by the content retrieved from the knowledge base.\n2. **No fabrication**: do not generate, infer, or create information that is not explicitly present in the knowledge base.\n3. **Accuracy first**: prefer incompleteness over inaccurate content.\n4. **Output format**:\n   * Hierarchically clear modular structure\n   * Logical grouping according to the MECE principle\n   * Professionally presented formatting\n   * Step-by-step cognitive guidance\n   * Reasonable use of headings and dividers for clarity\n   * *Italicize* key parameters\n   * **Bold** critical information\n5. **LaTeX formula requirements**:\n   * Inline formulas: start and end with `$`\n   * Block formulas: start and end with `$$`, each `$$` on its own line\n   * Block formula content must comply with LaTeX math syntax\n   * Verify formula correctness\n---\n## Additional Notes (Interaction & Failure Strategy)\n* If the knowledge base does not cover critical facts: explicitly inform the user (with sample wording)\n* For time-sensitive issues: enforce time filtering in the search request, and indicate the latest retrieval date in the answer.\n* Language requirement: answer in the user\u2019s preferred language\n",
-                            "temperature": "0.1",
-                            "temperatureEnabled": true,
-                            "tools": [
-                                {
-                                    "component_name": "Retrieval",
-                                    "name": "Retrieval",
-                                    "params": {
-                                        "cross_languages": [],
-                                        "description": "",
-                                        "empty_response": "",
-                                        "kb_ids": [],
-                                        "keywords_similarity_weight": 0.7,
-                                        "outputs": {
-                                            "formalized_content": {
-                                                "type": "string",
-                                                "value": ""
-                                            }
-                                        },
-                                        "rerank_id": "",
-                                        "similarity_threshold": 0.2,
-                                        "top_k": 1024,
-                                        "top_n": 8,
-                                        "use_kg": false
-                                    }
-                                }
-                            ],
-                            "topPEnabled": false,
-                            "top_p": 0.75,
-                            "user_prompt": "",
-                            "visual_files_var": ""
-                        },
-                        "label": "Agent",
-                        "name": "Knowledge Base Agent"
-                    },
-                    "dragging": false,
-                    "id": "Agent:NewPumasLick",
-                    "measured": {
-                        "height": 84,
-                        "width": 200
-                    },
-                    "position": {
-                        "x": 347.00048227952215,
-                        "y": 186.49109364794631
-                    },
-                    "selected": false,
-                    "sourcePosition": "right",
-                    "targetPosition": "left",
-                    "type": "agentNode"
-                },
-                {
-                    "data": {
-                        "form": {
-                            "description": "This is an agent for a specific task.",
-                            "user_prompt": "This is the order you need to send to the agent."
-                        },
-                        "label": "Tool",
-                        "name": "flow.tool_10"
-                    },
-                    "dragging": false,
-                    "id": "Tool:AllBirdsNail",
-                    "measured": {
-                        "height": 48,
-                        "width": 200
-                    },
-                    "position": {
-                        "x": 220.24819746977118,
-                        "y": 403.31576836482583
-                    },
-                    "selected": false,
-                    "sourcePosition": "right",
-                    "targetPosition": "left",
-                    "type": "toolNode"
-                }
-            ]
-        },
-        "history": [],
-        "memory": [],
-        "messages": [],
-        "path": [],
-        "retrieval": []
-    },
-    "avatar": "data:image/png;base64,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"
-}

agent/templates/market_seo_article_writer.json

@@ -1,14 +1,15 @@
 {
     "id": 12,
     "title": {
-        "en": "Generate SEO Blog",
-        "de": "SEO Blog generieren",
-        "zh": "生成SEO博客"},
+        "en": "SEO article writer",
+        "de": "SEO-Blog-Magnetiseur",
+        "zh": "SEO 博客写手"},
     "description": {
-        "en": "This workflow automatically generates a complete SEO-optimized blog article based on a simple user input. You don't need any writing experience. Just provide a topic or short request — the system will handle the rest.",
-        "de": "Dieser Workflow generiert automatisch einen vollständigen SEO-optimierten Blogartikel basierend auf einer einfachen Benutzereingabe. Sie benötigen keine Schreiberfahrung. Geben Sie einfach ein Thema oder eine kurze Anfrage ein – das System übernimmt den Rest.",
-        "zh": "此工作流根据简单的用户输入自动生成完整的SEO博客文章。你无需任何写作经验，只需提供一个主题或简短请求，系统将处理其余部分。"},
+        "en": "This SEO article writer automatically generates a complete SEO-optimized blog article based on a simple user input. You don't need any writing experience. Just provide a topic or short request — the system will handle the rest.",
+        "de": "SEO-Blog-Magnetiseur automatisch einen vollständigen SEO-optimierten Blogartikel basierend auf einer einfachen Benutzereingabe. Sie benötigen keine Schreiberfahrung. Geben Sie einfach ein Thema oder eine kurze Anfrage ein – das System übernimmt den Rest.",
+        "zh": "此 SEO 博客写手根据简单的用户输入自动生成完整的SEO博客文章。你无需任何写作经验，只需提供一个主题或简短请求，系统将处理其余部分。"},
     "canvas_type": "Marketing",
+    "canvas_types": ["Marketing", "Recommended"],
     "dsl": {
             "components": {
                 "Agent:BetterSitesSend": {
@@ -918,4 +919,4 @@
             "retrieval": []
         },
     "avatar": "data:image/jpeg;base64,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"
-}
+}

agent/templates/photo_text_translator.json

@@ -1,13 +1,13 @@
 {
     "id": 13,
     "title": {
-        "en": "ImageLingo",
-        "de": "ImageLingo",
-        "zh": "图片解析"},
+        "en": "Photo text translator",
+        "de": "Bild-Dolmetscher",
+        "zh": "图片文字快译"},
     "description": {
-        "en": "ImageLingo lets you snap any photo containing text—menus, signs, or documents—and instantly recognize and translate it into your language of choice using advanced AI-powered translation technology.",
-        "de": "ImageLingo ermöglicht es Ihnen, jedes Foto mit Text – Menüs, Schilder oder Dokumente – zu fotografieren und es sofort in Ihre gewünschte Sprache zu erkennen und zu übersetzen, unter Verwendung fortschrittlicher KI-gestützter Übersetzungstechnologie.",
-        "zh": "多模态大模型允许您拍摄任何包含文本的照片——菜单、标志或文档——立即识别并转换成您选择的语言。"},
+        "en": "Photo text translator lets you snap any photo containing text—menus, signs, or documents—and instantly recognize and translate it into your language of choice using advanced AI-powered translation technology.",
+        "de": "Bild-Dolmetscher ermöglicht es Ihnen, jedes Foto mit Text – Menüs, Schilder oder Dokumente – zu fotografieren und es sofort in Ihre gewünschte Sprache zu erkennen und zu übersetzen, unter Verwendung fortschrittlicher KI-gestützter Übersetzungstechnologie.",
+        "zh": "图片文字快译允许您拍摄任何包含文本的照片——菜单、标志或文档——立即识别并转换成您选择的语言。"},
     "canvas_type": "Consumer App",
 	"dsl": {
 		"components": {

agent/templates/reflective_academic_paper_generator.json

@@ -1,14 +1,15 @@
 {
     "id": 20,
     "title": {
-        "en": "Report Agent Using Knowledge Base",
-        "de": "Berichtsagent mit Wissensdatenbank",
-        "zh": "知识库检索智能体"},
+        "en": "Reflective academic paper generator",
+        "de": "Schreibhilfe für Reflexionspapiere",
+        "zh": "学术论文生成助手"},
     "description": {
-        "en": "A report generation assistant using local knowledge base, with advanced capabilities in task planning, reasoning, and reflective analysis. Recommended for academic research paper Q&A",
+        "en": "A reflective academic paper generator using local knowledge base, with advanced capabilities in task planning, reasoning, and reflective analysis. Recommended for academic research paper Q&A",
         "de": "Ein Berichtsgenerierungsassistent, der eine lokale Wissensdatenbank nutzt, mit erweiterten Fähigkeiten in Aufgabenplanung, Schlussfolgerung und reflektierender Analyse. Empfohlen für akademische Forschungspapier-Fragen und -Antworten.",
-        "zh": "一个使用本地知识库的报告生成助手，具备高级能力，包括任务规划、推理和反思性分析。推荐用于学术研究论文问答。"},
+        "zh": "一个使用本地知识库的学术论文生成助手，具备高级能力，包括任务规划、推理和反思性分析。推荐用于学术研究论文问答。"},
     "canvas_type": "Agent",
+    "canvas_types": ["Agent", "Recommended"],
     "dsl": {
         "components": {
             "Agent:NewPumasLick": {
@@ -330,4 +331,4 @@
         "retrieval": []
     },
     "avatar": "data:image/png;base64,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"
-}
+}

agent/templates/seo_article_writer.json

@@ -1,13 +1,13 @@
 {
     "id": 8,
     "title": {
-        "en": "Generate SEO Blog",
-        "de": "SEO Blog generieren",
-        "zh": "生成SEO博客"},
+        "en": "SEO article writer",
+        "de": "SEO-Blog-Magnetiseur",
+        "zh": "SEO 博客写手"},
     "description": {
         "en": "This is a multi-agent version of the SEO blog generation workflow. It simulates a small team of AI “writers”, where each agent plays a specialized role — just like a real editorial team.",
         "de": "Dies ist eine Multi-Agenten-Version des Workflows zur Erstellung von SEO-Blogs. Sie simuliert ein kleines Team von KI-„Autoren“, in dem jeder Agent eine spezielle Rolle übernimmt – genau wie in einem echten Redaktionsteam.",
-        "zh": "多智能体架构可根据简单的用户输入自动生成完整的SEO博客文章。模拟小型“作家”团队，其中每个智能体扮演一个专业角色——就像真正的编辑团队。"},
+        "zh": "SEO 博客写手可根据简单的用户输入自动生成完整的SEO博客文章。模拟小型“作家”团队，其中每个智能体扮演一个专业角色——就像真正的编辑团队。"},
     "canvas_type": "Agent",
     "dsl": {
             "components": {

agent/templates/seo_blog.json

@@ -1,921 +0,0 @@
-{
-    "id": 4,
-    "title": {
-        "en": "Generate SEO Blog",
-        "de": "SEO Blog generieren",
-        "zh": "生成SEO博客"},
-    "description": {
-        "en": "This workflow automatically generates a complete SEO-optimized blog article based on a simple user input. You don't need any writing experience. Just provide a topic or short request — the system will handle the rest.",
-        "de": "Dieser Workflow generiert automatisch einen vollständigen SEO-optimierten Blogartikel basierend auf einer einfachen Benutzereingabe. Sie benötigen keine Schreiberfahrung. Geben Sie einfach ein Thema oder eine kurze Anfrage ein – das System übernimmt den Rest.",
-        "zh": "此工作流根据简单的用户输入自动生成完整的SEO博客文章。你无需任何写作经验，只需提供一个主题或简短请求，系统将处理其余部分。"},
-    "canvas_type": "Recommended",
-    "dsl": {
-            "components": {
-                "Agent:BetterSitesSend": {
-                    "downstream": [
-                        "Agent:EagerNailsRemain"
-                    ],
-                    "obj": {
-                        "component_name": "Agent",
-                        "params": {
-                            "delay_after_error": 1,
-                            "description": "",
-                            "exception_comment": "",
-                            "exception_default_value": "",
-                            "exception_goto": [],
-                            "exception_method": null,
-                            "frequencyPenaltyEnabled": false,
-                            "frequency_penalty": 0.3,
-                            "llm_id": "deepseek-chat@DeepSeek",
-                            "maxTokensEnabled": false,
-                            "max_retries": 3,
-                            "max_rounds": 3,
-                            "max_tokens": 4096,
-                            "mcp": [],
-                            "message_history_window_size": 12,
-                            "outputs": {
-                                "content": {
-                                    "type": "string",
-                                    "value": ""
-                                }
-                            },
-                            "parameter": "Balance",
-                            "presencePenaltyEnabled": false,
-                            "presence_penalty": 0.2,
-                            "prompts": [
-                                {
-                                    "content": "The parse and keyword agent output is {Agent:ClearRabbitsScream@content}",
-                                    "role": "user"
-                                }
-                            ],
-                            "sys_prompt": "# Role\n\nYou are the **Outline_Agent**, responsible for generating a clear and SEO-optimized blog outline based on the user's parsed writing intent and keyword strategy.\n\n# Tool Access:\n\n- You have access to a search tool called `Tavily Search`.\n\n- If you are unsure how to structure a section, you may call this tool to search for related blog outlines or content from Google.\n\n- Do not overuse it. Your job is to extract **structure**, not to write paragraphs.\n\n\n# Goals\n\n1. Create a well-structured outline with appropriate H2 and H3 headings.\n\n2. Ensure logical flow from introduction to conclusion.\n\n3. Assign 1\u20132 suggested long-tail keywords to each major section for SEO alignment.\n\n4. Make the structure suitable for downstream paragraph writing.\n\n\n\n\n#Note\n\n- Use concise, scannable section titles.\n\n- Do not write full paragraphs.\n\n- Prioritize clarity, logical progression, and SEO alignment.\n\n\n\n- If the blog type is \u201cTutorial\u201d or \u201cHow-to\u201d, include step-based sections.\n\n\n# Input\n\nYou will receive:\n\n- Writing Type (e.g., Tutorial, Informative Guide)\n\n- Target Audience\n\n- User Intent Summary\n\n- 3\u20135 long-tail keywords\n\n\nUse this information to design a structure that both informs readers and maximizes search engine visibility.\n\n# Output Format\n\n```markdown\n\n## Blog Title (suggested)\n\n[Give a short, SEO-friendly title suggestion]\n\n## Outline\n\n### Introduction\n\n- Purpose of the article\n\n- Brief context\n\n- **Suggested keywords**: [keyword1, keyword2]\n\n### H2: [Section Title 1]\n\n- [Short description of what this section will cover]\n\n- **Suggested keywords**: [keyword1, keyword2]\n\n### H2: [Section Title 2]\n\n- [Short description of what this section will cover]\n\n- **Suggested keywords**: [keyword1, keyword2]\n\n### H2: [Section Title 3]\n\n- [Optional H3 Subsection Title A]\n\n  - [Explanation of sub-point]\n\n- [Optional H3 Subsection Title B]\n\n  - [Explanation of sub-point]\n\n- **Suggested keywords**: [keyword1]\n\n### Conclusion\n\n- Recap key takeaways\n\n- Optional CTA (Call to Action)\n\n- **Suggested keywords**: [keyword3]\n\n",
-                            "temperature": 0.5,
-                            "temperatureEnabled": true,
-                            "tools": [
-                                {
-                                    "component_name": "TavilySearch",
-                                    "name": "TavilySearch",
-                                    "params": {
-                                        "api_key": "",
-                                        "days": 7,
-                                        "exclude_domains": [],
-                                        "include_answer": false,
-                                        "include_domains": [],
-                                        "include_image_descriptions": false,
-                                        "include_images": false,
-                                        "include_raw_content": true,
-                                        "max_results": 5,
-                                        "outputs": {
-                                            "formalized_content": {
-                                                "type": "string",
-                                                "value": ""
-                                            },
-                                            "json": {
-                                                "type": "Array<Object>",
-                                                "value": []
-                                            }
-                                        },
-                                        "query": "sys.query",
-                                        "search_depth": "basic",
-                                        "topic": "general"
-                                    }
-                                }
-                            ],
-                            "topPEnabled": false,
-                            "top_p": 0.85,
-                            "user_prompt": "",
-                            "visual_files_var": ""
-                        }
-                    },
-                    "upstream": [
-                        "Agent:ClearRabbitsScream"
-                    ]
-                },
-                "Agent:ClearRabbitsScream": {
-                    "downstream": [
-                        "Agent:BetterSitesSend"
-                    ],
-                    "obj": {
-                        "component_name": "Agent",
-                        "params": {
-                            "delay_after_error": 1,
-                            "description": "",
-                            "exception_comment": "",
-                            "exception_default_value": "",
-                            "exception_goto": [],
-                            "exception_method": null,
-                            "frequencyPenaltyEnabled": false,
-                            "frequency_penalty": 0.5,
-                            "llm_id": "deepseek-chat@DeepSeek",
-                            "maxTokensEnabled": false,
-                            "max_retries": 3,
-                            "max_rounds": 1,
-                            "max_tokens": 4096,
-                            "mcp": [],
-                            "message_history_window_size": 12,
-                            "outputs": {
-                                "content": {
-                                    "type": "string",
-                                    "value": ""
-                                }
-                            },
-                            "parameter": "Precise",
-                            "presencePenaltyEnabled": false,
-                            "presence_penalty": 0.5,
-                            "prompts": [
-                                {
-                                    "content": "The user query is {sys.query}",
-                                    "role": "user"
-                                }
-                            ],
-                            "sys_prompt": "# Role\n\nYou are the **Parse_And_Keyword_Agent**, responsible for interpreting a user's blog writing request and generating a structured writing intent summary and keyword strategy for SEO-optimized content generation.\n\n# Goals\n\n1. Extract and infer the user's true writing intent, even if the input is informal or vague.\n\n2. Identify the writing type, target audience, and implied goal.\n\n3. Suggest 3\u20135 long-tail keywords based on the input and context.\n\n4. Output all data in a Markdown format for downstream agents.\n\n# Operating Guidelines\n\n\n- If the user's input lacks clarity, make reasonable and **conservative** assumptions based on SEO best practices.\n\n- Always choose one clear \"Writing Type\" from the list below.\n\n- Your job is not to write the blog \u2014 only to structure the brief.\n\n# Output Format\n\n```markdown\n## Writing Type\n\n[Choose one: Tutorial / Informative Guide / Marketing Content / Case Study / Opinion Piece / How-to / Comparison Article]\n\n## Target Audience\n\n[Try to be specific based on clues in the input: e.g., marketing managers, junior developers, SEO beginners]\n\n## User Intent Summary\n\n[A 1\u20132 sentence summary of what the user wants to achieve with the blog post]\n\n## Suggested Long-tail Keywords\n\n- keyword 1\n\n- keyword 2\n\n- keyword 3\n\n- keyword 4 (optional)\n\n- keyword 5 (optional)\n\n\n\n\n## Input Examples (and how to handle them)\n\nInput: \"I want to write about RAGFlow.\"\n\u2192 Output: Informative Guide, Audience: AI developers, Intent: explain what RAGFlow is and its use cases\n\nInput: \"Need a blog to promote our prompt design tool.\"\n\u2192 Output: Marketing Content, Audience: product managers or tool adopters, Intent: raise awareness and interest in the product\n\n\n\nInput: \"How to get more Google traffic using AI\"\n\u2192 Output: How-to, Audience: SEO marketers, Intent: guide readers on applying AI for SEO growth",
-                            "temperature": 0.2,
-                            "temperatureEnabled": true,
-                            "tools": [],
-                            "topPEnabled": false,
-                            "top_p": 0.75,
-                            "user_prompt": "",
-                            "visual_files_var": ""
-                        }
-                    },
-                    "upstream": [
-                        "begin"
-                    ]
-                },
-                "Agent:EagerNailsRemain": {
-                    "downstream": [
-                        "Agent:LovelyHeadsOwn"
-                    ],
-                    "obj": {
-                        "component_name": "Agent",
-                        "params": {
-                            "delay_after_error": 1,
-                            "description": "",
-                            "exception_comment": "",
-                            "exception_default_value": "",
-                            "exception_goto": [],
-                            "exception_method": null,
-                            "frequencyPenaltyEnabled": false,
-                            "frequency_penalty": 0.5,
-                            "llm_id": "deepseek-chat@DeepSeek",
-                            "maxTokensEnabled": false,
-                            "max_retries": 3,
-                            "max_rounds": 5,
-                            "max_tokens": 4096,
-                            "mcp": [],
-                            "message_history_window_size": 12,
-                            "outputs": {
-                                "content": {
-                                    "type": "string",
-                                    "value": ""
-                                }
-                            },
-                            "parameter": "Precise",
-                            "presencePenaltyEnabled": false,
-                            "presence_penalty": 0.5,
-                            "prompts": [
-                                {
-                                    "content": "The parse and keyword agent output is {Agent:ClearRabbitsScream@content}\n\n\n\nThe Outline agent output is {Agent:BetterSitesSend@content}",
-                                    "role": "user"
-                                }
-                            ],
-                            "sys_prompt": "# Role\n\nYou are the **Body_Agent**, responsible for generating the full content of each section of an SEO-optimized blog based on the provided outline and keyword strategy.\n\n# Tool Access:\n\nYou can use the `Tavily Search` tool to retrieve relevant content, statistics, or examples to support each section you're writing.\n\nUse it **only** when the provided outline lacks enough information, or if the section requires factual grounding.\n\nAlways cite the original link or indicate source where possible.\n\n\n# Goals\n\n1. Write each section (based on H2/H3 structure) as a complete and natural blog paragraph.\n\n2. Integrate the suggested long-tail keywords naturally into each section.\n\n3. When appropriate, use the `Tavily Search` tool to enrich your writing with relevant facts, examples, or quotes.\n\n4. Ensure each section is clear, engaging, and informative, suitable for both human readers and search engines.\n\n\n# Style Guidelines\n\n- Write in a tone appropriate to the audience. Be explanatory, not promotional, unless it's a marketing blog.\n\n- Avoid generic filler content. Prioritize clarity, structure, and value.\n\n- Ensure SEO keywords are embedded seamlessly, not forcefully.\n\n\n\n- Maintain writing rhythm. Vary sentence lengths. Use transitions between ideas.\n\n\n# Input\n\n\nYou will receive:\n\n- Blog title\n\n- Structured outline (including section titles, keywords, and descriptions)\n\n- Target audience\n\n- Blog type and user intent\n\nYou must **follow the outline strictly**. Write content **section-by-section**, based on the structure.\n\n\n# Output Format\n\n```markdown\n\n## H2: [Section Title]\n\n[Your generated content for this section \u2014 500-600 words, using keywords naturally.]\n\n",
-                            "temperature": 0.2,
-                            "temperatureEnabled": true,
-                            "tools": [
-                                {
-                                    "component_name": "TavilySearch",
-                                    "name": "TavilySearch",
-                                    "params": {
-                                        "api_key": "",
-                                        "days": 7,
-                                        "exclude_domains": [],
-                                        "include_answer": false,
-                                        "include_domains": [],
-                                        "include_image_descriptions": false,
-                                        "include_images": false,
-                                        "include_raw_content": true,
-                                        "max_results": 5,
-                                        "outputs": {
-                                            "formalized_content": {
-                                                "type": "string",
-                                                "value": ""
-                                            },
-                                            "json": {
-                                                "type": "Array<Object>",
-                                                "value": []
-                                            }
-                                        },
-                                        "query": "sys.query",
-                                        "search_depth": "basic",
-                                        "topic": "general"
-                                    }
-                                }
-                            ],
-                            "topPEnabled": false,
-                            "top_p": 0.75,
-                            "user_prompt": "",
-                            "visual_files_var": ""
-                        }
-                    },
-                    "upstream": [
-                        "Agent:BetterSitesSend"
-                    ]
-                },
-                "Agent:LovelyHeadsOwn": {
-                    "downstream": [
-                        "Message:LegalBeansBet"
-                    ],
-                    "obj": {
-                        "component_name": "Agent",
-                        "params": {
-                            "delay_after_error": 1,
-                            "description": "",
-                            "exception_comment": "",
-                            "exception_default_value": "",
-                            "exception_goto": [],
-                            "exception_method": null,
-                            "frequencyPenaltyEnabled": false,
-                            "frequency_penalty": 0.5,
-                            "llm_id": "deepseek-chat@DeepSeek",
-                            "maxTokensEnabled": false,
-                            "max_retries": 3,
-                            "max_rounds": 5,
-                            "max_tokens": 4096,
-                            "mcp": [],
-                            "message_history_window_size": 12,
-                            "outputs": {
-                                "content": {
-                                    "type": "string",
-                                    "value": ""
-                                }
-                            },
-                            "parameter": "Precise",
-                            "presencePenaltyEnabled": false,
-                            "presence_penalty": 0.5,
-                            "prompts": [
-                                {
-                                    "content": "The parse and keyword agent output is {Agent:ClearRabbitsScream@content}\n\nThe Outline agent output is {Agent:BetterSitesSend@content}\n\nThe Body agent output is {Agent:EagerNailsRemain@content}",
-                                    "role": "user"
-                                }
-                            ],
-                            "sys_prompt": "# Role\n\nYou are the **Editor_Agent**, responsible for finalizing the blog post for both human readability and SEO effectiveness.\n\n# Goals\n\n1. Polish the entire blog content for clarity, coherence, and style.\n\n2. Improve transitions between sections, ensure logical flow.\n\n3. Verify that keywords are used appropriately and effectively.\n\n4. Conduct a lightweight SEO audit \u2014 checking keyword density, structure (H1/H2/H3), and overall searchability.\n\n\n\n# Style Guidelines\n\n- Be precise. Avoid bloated or vague language.\n\n- Maintain an informative and engaging tone, suitable to the target audience.\n\n- Do not remove keywords unless absolutely necessary for clarity.\n\n- Ensure paragraph flow and section continuity.\n\n\n# Input\n\nYou will receive:\n\n- Full blog content, written section-by-section\n\n- Original outline with suggested keywords\n\n- Target audience and writing type\n\n# Output Format\n\n```markdown\n\n[The revised, fully polished blog post content goes here.]\n\n",
-                            "temperature": 0.2,
-                            "temperatureEnabled": true,
-                            "tools": [],
-                            "topPEnabled": false,
-                            "top_p": 0.75,
-                            "user_prompt": "",
-                            "visual_files_var": ""
-                        }
-                    },
-                    "upstream": [
-                        "Agent:EagerNailsRemain"
-                    ]
-                },
-                "Message:LegalBeansBet": {
-                    "downstream": [],
-                    "obj": {
-                        "component_name": "Message",
-                        "params": {
-                            "content": [
-                                "{Agent:LovelyHeadsOwn@content}"
-                            ]
-                        }
-                    },
-                    "upstream": [
-                        "Agent:LovelyHeadsOwn"
-                    ]
-                },
-                "begin": {
-                    "downstream": [
-                        "Agent:ClearRabbitsScream"
-                    ],
-                    "obj": {
-                        "component_name": "Begin",
-                        "params": {
-                            "enablePrologue": true,
-                            "inputs": {},
-                            "mode": "conversational",
-                            "prologue": "Hi! I'm your SEO blog assistant.\n\nTo get started, please tell me:\n1. What topic you want the blog to cover\n2. Who is the target audience\n3. What you hope to achieve with this blog (e.g., SEO traffic, teaching beginners, promoting a product)\n"
-                        }
-                    },
-                    "upstream": []
-                }
-            },
-            "globals": {
-                "sys.conversation_turns": 0,
-                "sys.files": [],
-                "sys.query": "",
-                "sys.user_id": ""
-            },
-            "graph": {
-                "edges": [
-                    {
-                        "data": {
-                            "isHovered": false
-                        },
-                        "id": "xy-edge__beginstart-Agent:ClearRabbitsScreamend",
-                        "source": "begin",
-                        "sourceHandle": "start",
-                        "target": "Agent:ClearRabbitsScream",
-                        "targetHandle": "end"
-                    },
-                    {
-                        "data": {
-                            "isHovered": false
-                        },
-                        "id": "xy-edge__Agent:ClearRabbitsScreamstart-Agent:BetterSitesSendend",
-                        "source": "Agent:ClearRabbitsScream",
-                        "sourceHandle": "start",
-                        "target": "Agent:BetterSitesSend",
-                        "targetHandle": "end"
-                    },
-                    {
-                        "data": {
-                            "isHovered": false
-                        },
-                        "id": "xy-edge__Agent:BetterSitesSendtool-Tool:SharpPensBurnend",
-                        "source": "Agent:BetterSitesSend",
-                        "sourceHandle": "tool",
-                        "target": "Tool:SharpPensBurn",
-                        "targetHandle": "end"
-                    },
-                    {
-                        "data": {
-                            "isHovered": false
-                        },
-                        "id": "xy-edge__Agent:BetterSitesSendstart-Agent:EagerNailsRemainend",
-                        "source": "Agent:BetterSitesSend",
-                        "sourceHandle": "start",
-                        "target": "Agent:EagerNailsRemain",
-                        "targetHandle": "end"
-                    },
-                    {
-                        "id": "xy-edge__Agent:EagerNailsRemaintool-Tool:WickedDeerHealend",
-                        "source": "Agent:EagerNailsRemain",
-                        "sourceHandle": "tool",
-                        "target": "Tool:WickedDeerHeal",
-                        "targetHandle": "end"
-                    },
-                    {
-                        "data": {
-                            "isHovered": false
-                        },
-                        "id": "xy-edge__Agent:EagerNailsRemainstart-Agent:LovelyHeadsOwnend",
-                        "source": "Agent:EagerNailsRemain",
-                        "sourceHandle": "start",
-                        "target": "Agent:LovelyHeadsOwn",
-                        "targetHandle": "end"
-                    },
-                    {
-                        "data": {
-                            "isHovered": false
-                        },
-                        "id": "xy-edge__Agent:LovelyHeadsOwnstart-Message:LegalBeansBetend",
-                        "source": "Agent:LovelyHeadsOwn",
-                        "sourceHandle": "start",
-                        "target": "Message:LegalBeansBet",
-                        "targetHandle": "end"
-                    }
-                ],
-                "nodes": [
-                    {
-                        "data": {
-                            "form": {
-                                "enablePrologue": true,
-                                "inputs": {},
-                                "mode": "conversational",
-                                "prologue": "Hi! I'm your SEO blog assistant.\n\nTo get started, please tell me:\n1. What topic you want the blog to cover\n2. Who is the target audience\n3. What you hope to achieve with this blog (e.g., SEO traffic, teaching beginners, promoting a product)\n"
-                            },
-                            "label": "Begin",
-                            "name": "begin"
-                        },
-                        "id": "begin",
-                        "measured": {
-                            "height": 48,
-                            "width": 200
-                        },
-                        "position": {
-                            "x": 50,
-                            "y": 200
-                        },
-                        "selected": false,
-                        "sourcePosition": "left",
-                        "targetPosition": "right",
-                        "type": "beginNode"
-                    },
-                    {
-                        "data": {
-                            "form": {
-                                "delay_after_error": 1,
-                                "description": "",
-                                "exception_comment": "",
-                                "exception_default_value": "",
-                                "exception_goto": [],
-                                "exception_method": null,
-                                "frequencyPenaltyEnabled": false,
-                                "frequency_penalty": 0.5,
-                                "llm_id": "deepseek-chat@DeepSeek",
-                                "maxTokensEnabled": false,
-                                "max_retries": 3,
-                                "max_rounds": 1,
-                                "max_tokens": 4096,
-                                "mcp": [],
-                                "message_history_window_size": 12,
-                                "outputs": {
-                                    "content": {
-                                        "type": "string",
-                                        "value": ""
-                                    }
-                                },
-                                "parameter": "Precise",
-                                "presencePenaltyEnabled": false,
-                                "presence_penalty": 0.5,
-                                "prompts": [
-                                    {
-                                        "content": "The user query is {sys.query}",
-                                        "role": "user"
-                                    }
-                                ],
-                                "sys_prompt": "# Role\n\nYou are the **Parse_And_Keyword_Agent**, responsible for interpreting a user's blog writing request and generating a structured writing intent summary and keyword strategy for SEO-optimized content generation.\n\n# Goals\n\n1. Extract and infer the user's true writing intent, even if the input is informal or vague.\n\n2. Identify the writing type, target audience, and implied goal.\n\n3. Suggest 3\u20135 long-tail keywords based on the input and context.\n\n4. Output all data in a Markdown format for downstream agents.\n\n# Operating Guidelines\n\n\n- If the user's input lacks clarity, make reasonable and **conservative** assumptions based on SEO best practices.\n\n- Always choose one clear \"Writing Type\" from the list below.\n\n- Your job is not to write the blog \u2014 only to structure the brief.\n\n# Output Format\n\n```markdown\n## Writing Type\n\n[Choose one: Tutorial / Informative Guide / Marketing Content / Case Study / Opinion Piece / How-to / Comparison Article]\n\n## Target Audience\n\n[Try to be specific based on clues in the input: e.g., marketing managers, junior developers, SEO beginners]\n\n## User Intent Summary\n\n[A 1\u20132 sentence summary of what the user wants to achieve with the blog post]\n\n## Suggested Long-tail Keywords\n\n- keyword 1\n\n- keyword 2\n\n- keyword 3\n\n- keyword 4 (optional)\n\n- keyword 5 (optional)\n\n\n\n\n## Input Examples (and how to handle them)\n\nInput: \"I want to write about RAGFlow.\"\n\u2192 Output: Informative Guide, Audience: AI developers, Intent: explain what RAGFlow is and its use cases\n\nInput: \"Need a blog to promote our prompt design tool.\"\n\u2192 Output: Marketing Content, Audience: product managers or tool adopters, Intent: raise awareness and interest in the product\n\n\n\nInput: \"How to get more Google traffic using AI\"\n\u2192 Output: How-to, Audience: SEO marketers, Intent: guide readers on applying AI for SEO growth",
-                                "temperature": 0.2,
-                                "temperatureEnabled": true,
-                                "tools": [],
-                                "topPEnabled": false,
-                                "top_p": 0.75,
-                                "user_prompt": "",
-                                "visual_files_var": ""
-                            },
-                            "label": "Agent",
-                            "name": "Parse And Keyword Agent"
-                        },
-                        "dragging": false,
-                        "id": "Agent:ClearRabbitsScream",
-                        "measured": {
-                            "height": 84,
-                            "width": 200
-                        },
-                        "position": {
-                            "x": 344.7766966202233,
-                            "y": 234.82202253184496
-                        },
-                        "selected": false,
-                        "sourcePosition": "right",
-                        "targetPosition": "left",
-                        "type": "agentNode"
-                    },
-                    {
-                        "data": {
-                            "form": {
-                                "delay_after_error": 1,
-                                "description": "",
-                                "exception_comment": "",
-                                "exception_default_value": "",
-                                "exception_goto": [],
-                                "exception_method": null,
-                                "frequencyPenaltyEnabled": false,
-                                "frequency_penalty": 0.3,
-                                "llm_id": "deepseek-chat@DeepSeek",
-                                "maxTokensEnabled": false,
-                                "max_retries": 3,
-                                "max_rounds": 3,
-                                "max_tokens": 4096,
-                                "mcp": [],
-                                "message_history_window_size": 12,
-                                "outputs": {
-                                    "content": {
-                                        "type": "string",
-                                        "value": ""
-                                    }
-                                },
-                                "parameter": "Balance",
-                                "presencePenaltyEnabled": false,
-                                "presence_penalty": 0.2,
-                                "prompts": [
-                                    {
-                                        "content": "The parse and keyword agent output is {Agent:ClearRabbitsScream@content}",
-                                        "role": "user"
-                                    }
-                                ],
-                                "sys_prompt": "# Role\n\nYou are the **Outline_Agent**, responsible for generating a clear and SEO-optimized blog outline based on the user's parsed writing intent and keyword strategy.\n\n# Tool Access:\n\n- You have access to a search tool called `Tavily Search`.\n\n- If you are unsure how to structure a section, you may call this tool to search for related blog outlines or content from Google.\n\n- Do not overuse it. Your job is to extract **structure**, not to write paragraphs.\n\n\n# Goals\n\n1. Create a well-structured outline with appropriate H2 and H3 headings.\n\n2. Ensure logical flow from introduction to conclusion.\n\n3. Assign 1\u20132 suggested long-tail keywords to each major section for SEO alignment.\n\n4. Make the structure suitable for downstream paragraph writing.\n\n\n\n\n#Note\n\n- Use concise, scannable section titles.\n\n- Do not write full paragraphs.\n\n- Prioritize clarity, logical progression, and SEO alignment.\n\n\n\n- If the blog type is \u201cTutorial\u201d or \u201cHow-to\u201d, include step-based sections.\n\n\n# Input\n\nYou will receive:\n\n- Writing Type (e.g., Tutorial, Informative Guide)\n\n- Target Audience\n\n- User Intent Summary\n\n- 3\u20135 long-tail keywords\n\n\nUse this information to design a structure that both informs readers and maximizes search engine visibility.\n\n# Output Format\n\n```markdown\n\n## Blog Title (suggested)\n\n[Give a short, SEO-friendly title suggestion]\n\n## Outline\n\n### Introduction\n\n- Purpose of the article\n\n- Brief context\n\n- **Suggested keywords**: [keyword1, keyword2]\n\n### H2: [Section Title 1]\n\n- [Short description of what this section will cover]\n\n- **Suggested keywords**: [keyword1, keyword2]\n\n### H2: [Section Title 2]\n\n- [Short description of what this section will cover]\n\n- **Suggested keywords**: [keyword1, keyword2]\n\n### H2: [Section Title 3]\n\n- [Optional H3 Subsection Title A]\n\n  - [Explanation of sub-point]\n\n- [Optional H3 Subsection Title B]\n\n  - [Explanation of sub-point]\n\n- **Suggested keywords**: [keyword1]\n\n### Conclusion\n\n- Recap key takeaways\n\n- Optional CTA (Call to Action)\n\n- **Suggested keywords**: [keyword3]\n\n",
-                                "temperature": 0.5,
-                                "temperatureEnabled": true,
-                                "tools": [
-                                    {
-                                        "component_name": "TavilySearch",
-                                        "name": "TavilySearch",
-                                        "params": {
-                                            "api_key": "",
-                                            "days": 7,
-                                            "exclude_domains": [],
-                                            "include_answer": false,
-                                            "include_domains": [],
-                                            "include_image_descriptions": false,
-                                            "include_images": false,
-                                            "include_raw_content": true,
-                                            "max_results": 5,
-                                            "outputs": {
-                                                "formalized_content": {
-                                                    "type": "string",
-                                                    "value": ""
-                                                },
-                                                "json": {
-                                                    "type": "Array<Object>",
-                                                    "value": []
-                                                }
-                                            },
-                                            "query": "sys.query",
-                                            "search_depth": "basic",
-                                            "topic": "general"
-                                        }
-                                    }
-                                ],
-                                "topPEnabled": false,
-                                "top_p": 0.85,
-                                "user_prompt": "",
-                                "visual_files_var": ""
-                            },
-                            "label": "Agent",
-                            "name": "Outline Agent"
-                        },
-                        "dragging": false,
-                        "id": "Agent:BetterSitesSend",
-                        "measured": {
-                            "height": 84,
-                            "width": 200
-                        },
-                        "position": {
-                            "x": 613.4368763415628,
-                            "y": 164.3074269048589
-                        },
-                        "selected": false,
-                        "sourcePosition": "right",
-                        "targetPosition": "left",
-                        "type": "agentNode"
-                    },
-                    {
-                        "data": {
-                            "form": {
-                                "description": "This is an agent for a specific task.",
-                                "user_prompt": "This is the order you need to send to the agent."
-                            },
-                            "label": "Tool",
-                            "name": "flow.tool_0"
-                        },
-                        "dragging": false,
-                        "id": "Tool:SharpPensBurn",
-                        "measured": {
-                            "height": 44,
-                            "width": 200
-                        },
-                        "position": {
-                            "x": 580.1877078861457,
-                            "y": 287.7669662022325
-                        },
-                        "selected": false,
-                        "sourcePosition": "right",
-                        "targetPosition": "left",
-                        "type": "toolNode"
-                    },
-                    {
-                        "data": {
-                            "form": {
-                                "delay_after_error": 1,
-                                "description": "",
-                                "exception_comment": "",
-                                "exception_default_value": "",
-                                "exception_goto": [],
-                                "exception_method": null,
-                                "frequencyPenaltyEnabled": false,
-                                "frequency_penalty": 0.5,
-                                "llm_id": "deepseek-chat@DeepSeek",
-                                "maxTokensEnabled": false,
-                                "max_retries": 3,
-                                "max_rounds": 5,
-                                "max_tokens": 4096,
-                                "mcp": [],
-                                "message_history_window_size": 12,
-                                "outputs": {
-                                    "content": {
-                                        "type": "string",
-                                        "value": ""
-                                    }
-                                },
-                                "parameter": "Precise",
-                                "presencePenaltyEnabled": false,
-                                "presence_penalty": 0.5,
-                                "prompts": [
-                                    {
-                                        "content": "The parse and keyword agent output is {Agent:ClearRabbitsScream@content}\n\n\n\nThe Outline agent output is {Agent:BetterSitesSend@content}",
-                                        "role": "user"
-                                    }
-                                ],
-                                "sys_prompt": "# Role\n\nYou are the **Body_Agent**, responsible for generating the full content of each section of an SEO-optimized blog based on the provided outline and keyword strategy.\n\n# Tool Access:\n\nYou can use the `Tavily Search` tool to retrieve relevant content, statistics, or examples to support each section you're writing.\n\nUse it **only** when the provided outline lacks enough information, or if the section requires factual grounding.\n\nAlways cite the original link or indicate source where possible.\n\n\n# Goals\n\n1. Write each section (based on H2/H3 structure) as a complete and natural blog paragraph.\n\n2. Integrate the suggested long-tail keywords naturally into each section.\n\n3. When appropriate, use the `Tavily Search` tool to enrich your writing with relevant facts, examples, or quotes.\n\n4. Ensure each section is clear, engaging, and informative, suitable for both human readers and search engines.\n\n\n# Style Guidelines\n\n- Write in a tone appropriate to the audience. Be explanatory, not promotional, unless it's a marketing blog.\n\n- Avoid generic filler content. Prioritize clarity, structure, and value.\n\n- Ensure SEO keywords are embedded seamlessly, not forcefully.\n\n\n\n- Maintain writing rhythm. Vary sentence lengths. Use transitions between ideas.\n\n\n# Input\n\n\nYou will receive:\n\n- Blog title\n\n- Structured outline (including section titles, keywords, and descriptions)\n\n- Target audience\n\n- Blog type and user intent\n\nYou must **follow the outline strictly**. Write content **section-by-section**, based on the structure.\n\n\n# Output Format\n\n```markdown\n\n## H2: [Section Title]\n\n[Your generated content for this section \u2014 500-600 words, using keywords naturally.]\n\n",
-                                "temperature": 0.2,
-                                "temperatureEnabled": true,
-                                "tools": [
-                                    {
-                                        "component_name": "TavilySearch",
-                                        "name": "TavilySearch",
-                                        "params": {
-                                            "api_key": "",
-                                            "days": 7,
-                                            "exclude_domains": [],
-                                            "include_answer": false,
-                                            "include_domains": [],
-                                            "include_image_descriptions": false,
-                                            "include_images": false,
-                                            "include_raw_content": true,
-                                            "max_results": 5,
-                                            "outputs": {
-                                                "formalized_content": {
-                                                    "type": "string",
-                                                    "value": ""
-                                                },
-                                                "json": {
-                                                    "type": "Array<Object>",
-                                                    "value": []
-                                                }
-                                            },
-                                            "query": "sys.query",
-                                            "search_depth": "basic",
-                                            "topic": "general"
-                                        }
-                                    }
-                                ],
-                                "topPEnabled": false,
-                                "top_p": 0.75,
-                                "user_prompt": "",
-                                "visual_files_var": ""
-                            },
-                            "label": "Agent",
-                            "name": "Body Agent"
-                        },
-                        "dragging": false,
-                        "id": "Agent:EagerNailsRemain",
-                        "measured": {
-                            "height": 84,
-                            "width": 200
-                        },
-                        "position": {
-                            "x": 889.0614605692713,
-                            "y": 247.00973041799065
-                        },
-                        "selected": false,
-                        "sourcePosition": "right",
-                        "targetPosition": "left",
-                        "type": "agentNode"
-                    },
-                    {
-                        "data": {
-                            "form": {
-                                "description": "This is an agent for a specific task.",
-                                "user_prompt": "This is the order you need to send to the agent."
-                            },
-                            "label": "Tool",
-                            "name": "flow.tool_1"
-                        },
-                        "dragging": false,
-                        "id": "Tool:WickedDeerHeal",
-                        "measured": {
-                            "height": 44,
-                            "width": 200
-                        },
-                        "position": {
-                            "x": 853.2006404239659,
-                            "y": 364.37541577229143
-                        },
-                        "selected": false,
-                        "sourcePosition": "right",
-                        "targetPosition": "left",
-                        "type": "toolNode"
-                    },
-                    {
-                        "data": {
-                            "form": {
-                                "delay_after_error": 1,
-                                "description": "",
-                                "exception_comment": "",
-                                "exception_default_value": "",
-                                "exception_goto": [],
-                                "exception_method": null,
-                                "frequencyPenaltyEnabled": false,
-                                "frequency_penalty": 0.5,
-                                "llm_id": "deepseek-chat@DeepSeek",
-                                "maxTokensEnabled": false,
-                                "max_retries": 3,
-                                "max_rounds": 5,
-                                "max_tokens": 4096,
-                                "mcp": [],
-                                "message_history_window_size": 12,
-                                "outputs": {
-                                    "content": {
-                                        "type": "string",
-                                        "value": ""
-                                    }
-                                },
-                                "parameter": "Precise",
-                                "presencePenaltyEnabled": false,
-                                "presence_penalty": 0.5,
-                                "prompts": [
-                                    {
-                                        "content": "The parse and keyword agent output is {Agent:ClearRabbitsScream@content}\n\nThe Outline agent output is {Agent:BetterSitesSend@content}\n\nThe Body agent output is {Agent:EagerNailsRemain@content}",
-                                        "role": "user"
-                                    }
-                                ],
-                                "sys_prompt": "# Role\n\nYou are the **Editor_Agent**, responsible for finalizing the blog post for both human readability and SEO effectiveness.\n\n# Goals\n\n1. Polish the entire blog content for clarity, coherence, and style.\n\n2. Improve transitions between sections, ensure logical flow.\n\n3. Verify that keywords are used appropriately and effectively.\n\n4. Conduct a lightweight SEO audit \u2014 checking keyword density, structure (H1/H2/H3), and overall searchability.\n\n\n\n# Style Guidelines\n\n- Be precise. Avoid bloated or vague language.\n\n- Maintain an informative and engaging tone, suitable to the target audience.\n\n- Do not remove keywords unless absolutely necessary for clarity.\n\n- Ensure paragraph flow and section continuity.\n\n\n# Input\n\nYou will receive:\n\n- Full blog content, written section-by-section\n\n- Original outline with suggested keywords\n\n- Target audience and writing type\n\n# Output Format\n\n```markdown\n\n[The revised, fully polished blog post content goes here.]\n\n",
-                                "temperature": 0.2,
-                                "temperatureEnabled": true,
-                                "tools": [],
-                                "topPEnabled": false,
-                                "top_p": 0.75,
-                                "user_prompt": "",
-                                "visual_files_var": ""
-                            },
-                            "label": "Agent",
-                            "name": "Editor Agent"
-                        },
-                        "dragging": false,
-                        "id": "Agent:LovelyHeadsOwn",
-                        "measured": {
-                            "height": 84,
-                            "width": 200
-                        },
-                        "position": {
-                            "x": 1160.3332919804993,
-                            "y": 149.50806732882472
-                        },
-                        "selected": false,
-                        "sourcePosition": "right",
-                        "targetPosition": "left",
-                        "type": "agentNode"
-                    },
-                    {
-                        "data": {
-                            "form": {
-                                "content": [
-                                    "{Agent:LovelyHeadsOwn@content}"
-                                ]
-                            },
-                            "label": "Message",
-                            "name": "Response"
-                        },
-                        "dragging": false,
-                        "id": "Message:LegalBeansBet",
-                        "measured": {
-                            "height": 56,
-                            "width": 200
-                        },
-                        "position": {
-                            "x": 1370.6665839609984,
-                            "y": 267.0323933738015
-                        },
-                        "selected": false,
-                        "sourcePosition": "right",
-                        "targetPosition": "left",
-                        "type": "messageNode"
-                    },
-                    {
-                        "data": {
-                            "form": {
-                                "text": "This workflow automatically generates a complete SEO-optimized blog article based on a simple user input. You don\u2019t need any writing experience. Just provide a topic or short request \u2014 the system will handle the rest.\n\nThe process includes the following key stages:\n\n1. **Understanding your topic and goals**\n2. **Designing the blog structure**\n3. **Writing high-quality content**\n\n\n"
-                            },
-                            "label": "Note",
-                            "name": "Workflow Overall Description"
-                        },
-                        "dragHandle": ".note-drag-handle",
-                        "dragging": false,
-                        "height": 205,
-                        "id": "Note:SlimyGhostsWear",
-                        "measured": {
-                            "height": 205,
-                            "width": 415
-                        },
-                        "position": {
-                            "x": -284.3143151688742,
-                            "y": 150.47632147913419
-                        },
-                        "resizing": false,
-                        "selected": false,
-                        "sourcePosition": "right",
-                        "targetPosition": "left",
-                        "type": "noteNode",
-                        "width": 415
-                    },
-                    {
-                        "data": {
-                            "form": {
-                                "text": "**Purpose**:  \nThis agent reads the user\u2019s input and figures out what kind of blog needs to be written.\n\n**What it does**:\n- Understands the main topic you want to write about  \n- Identifies who the blog is for (e.g., beginners, marketers, developers)  \n- Determines the writing purpose (e.g., SEO traffic, product promotion, education)  \n- Suggests 3\u20135 long-tail SEO keywords related to the topic"
-                            },
-                            "label": "Note",
-                            "name": "Parse And Keyword Agent"
-                        },
-                        "dragHandle": ".note-drag-handle",
-                        "dragging": false,
-                        "height": 152,
-                        "id": "Note:EmptyChairsShake",
-                        "measured": {
-                            "height": 152,
-                            "width": 340
-                        },
-                        "position": {
-                            "x": 295.04147626768133,
-                            "y": 372.2755718118446
-                        },
-                        "resizing": false,
-                        "selected": false,
-                        "sourcePosition": "right",
-                        "targetPosition": "left",
-                        "type": "noteNode",
-                        "width": 340
-                    },
-                    {
-                        "data": {
-                            "form": {
-                                "text": "**Purpose**:  \nThis agent builds the blog structure \u2014 just like writing a table of contents before you start writing the full article.\n\n**What it does**:\n- Suggests a clear blog title that includes important keywords  \n- Breaks the article into sections using H2 and H3 headings (like a professional blog layout)  \n- Assigns 1\u20132 recommended keywords to each section to help with SEO  \n- Follows the writing goal and target audience set in the previous step"
-                            },
-                            "label": "Note",
-                            "name": "Outline Agent"
-                        },
-                        "dragHandle": ".note-drag-handle",
-                        "dragging": false,
-                        "height": 146,
-                        "id": "Note:TallMelonsNotice",
-                        "measured": {
-                            "height": 146,
-                            "width": 343
-                        },
-                        "position": {
-                            "x": 598.5644991893463,
-                            "y": 5.801054564756448
-                        },
-                        "resizing": false,
-                        "selected": false,
-                        "sourcePosition": "right",
-                        "targetPosition": "left",
-                        "type": "noteNode",
-                        "width": 343
-                    },
-                    {
-                        "data": {
-                            "form": {
-                                "text": "**Purpose**:  \nThis agent is responsible for writing the actual content of the blog \u2014 paragraph by paragraph \u2014 based on the outline created earlier.\n\n**What it does**:\n- Looks at each H2/H3 section in the outline  \n- Writes 150\u2013220 words of clear, helpful, and well-structured content per section  \n- Includes the suggested SEO keywords naturally (not keyword stuffing)  \n- Uses real examples or facts if needed (by calling a web search tool like Tavily)"
-                            },
-                            "label": "Note",
-                            "name": "Body Agent"
-                        },
-                        "dragHandle": ".note-drag-handle",
-                        "dragging": false,
-                        "height": 137,
-                        "id": "Note:RipeCougarsBuild",
-                        "measured": {
-                            "height": 137,
-                            "width": 319
-                        },
-                        "position": {
-                            "x": 860.4854129814981,
-                            "y": 427.2196835690842
-                        },
-                        "resizing": false,
-                        "selected": false,
-                        "sourcePosition": "right",
-                        "targetPosition": "left",
-                        "type": "noteNode",
-                        "width": 319
-                    },
-                    {
-                        "data": {
-                            "form": {
-                                "text": "**Purpose**:  \nThis agent reviews the entire blog draft to make sure it is smooth, professional, and SEO-friendly. It acts like a human editor before publishing.\n\n**What it does**:\n- Polishes the writing: improves sentence clarity, fixes awkward phrasing  \n- Makes sure the content flows well from one section to the next  \n- Double-checks keyword usage: are they present, natural, and not overused?  \n- Verifies the blog structure (H1, H2, H3 headings) is correct  \n- Adds two key SEO elements:\n  - **Meta Title** (shows up in search results)\n  - **Meta Description** (summary for Google and social sharing)"
-                            },
-                            "label": "Note",
-                            "name": "Editor Agent"
-                        },
-                        "dragHandle": ".note-drag-handle",
-                        "height": 146,
-                        "id": "Note:OpenTurkeysSell",
-                        "measured": {
-                            "height": 146,
-                            "width": 320
-                        },
-                        "position": {
-                            "x": 1129,
-                            "y": -30
-                        },
-                        "resizing": false,
-                        "selected": false,
-                        "sourcePosition": "right",
-                        "targetPosition": "left",
-                        "type": "noteNode",
-                        "width": 320
-                    }
-                ]
-            },
-            "history": [],
-            "messages": [],
-            "path": [],
-            "retrieval": []
-        },
-    "avatar": "data:image/jpeg;base64,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"
-}

agent/templates/smart_customer_service_specialist.json

@@ -1,14 +1,14 @@
 {
     "id": 22,
     "title": {
-        "en": "Ecommerce Customer Service Workflow",
+        "en": "Smart customer service specialist",
         "de": "Ecommerce Kundenservice Workflow",
-        "zh": "电子商务客户服务工作流程"
+        "zh": "智能客户服务专员"
     },
     "description": {
-        "en": "This template helps e-commerce platforms address complex customer needs, such as comparing product features, providing usage support, and coordinating home installation services.",
-        "de": "Diese Vorlage hilft E-Commerce-Plattformen, komplexe Kundenbedürfnisse zu erfüllen, wie z.B. den Vergleich von Produktmerkmalen, die Bereitstellung von Nutzungsunterstützung und die Koordination von Hausinstallationsdiensten.",
-        "zh": "该模板可帮助电子商务平台解决复杂的客户需求，例如比较产品功能、提供使用支持和协调家庭安装服务。"
+        "en": "This template helps address complex customer needs, such as comparing product features, providing usage support, and coordinating home installation services.",
+        "de": "Diese Vorlage hilft komplexe Kundenbedürfnisse zu erfüllen, wie z.B. den Vergleich von Produktmerkmalen, die Bereitstellung von Nutzungsunterstützung und die Koordination von Hausinstallationsdiensten.",
+        "zh": "该模板可帮助解决复杂的客户需求，例如比较产品功能、提供使用支持和协调家庭安装服务。"
     },
     "canvas_type": "Customer Support",
     "dsl": {

agent/templates/text2sql_data_expert.json

@@ -1,12 +1,12 @@
 {
     "id": 17,
     "title": {
-        "en": "SQL Assistant",
-        "de": "SQL Assistent",
-        "zh": "SQL助理"},
+        "en": "Text-to-SQL data expert",
+        "de": "Text-to-SQL-Datenexperte",
+        "zh": "Text-to-SQL 问数专家"},
     "description": {
-        "en": "SQL Assistant is an AI-powered tool that lets business users turn plain-English questions into fully formed SQL queries. Simply type your question (e.g., 'Show me last quarter's top 10 products by revenue') and SQL Assistant generates the exact SQL, runs it against your database, and returns the results in seconds. ",
-        "de": "SQL-Assistent ist ein KI-gestütztes Tool, mit dem Geschäftsanwender einfache englische Fragen in vollständige SQL-Abfragen umwandeln können. Geben Sie einfach Ihre Frage ein (z.B. 'Zeige mir die Top 10 Produkte des letzten Quartals nach Umsatz') und der SQL-Assistent generiert das exakte SQL, führt es gegen Ihre Datenbank aus und liefert die Ergebnisse in Sekunden.",
+        "en": "Text-to-SQL data expert lets business users turn plain-English questions into fully formed SQL queries. Simply type your question (e.g., 'Show me last quarter's top 10 products by revenue') and Text-to-SQL data expert generates the exact SQL, runs it against your database, and returns the results in seconds. ",
+        "de": "Text-to-SQL-Datenexperte ist ein KI-gestütztes Tool, mit dem Geschäftsanwender einfache englische Fragen in vollständige SQL-Abfragen umwandeln können. Geben Sie einfach Ihre Frage ein (z.B. 'Zeige mir die Top 10 Produkte des letzten Quartals nach Umsatz') und der SQL-Assistent generiert das exakte SQL, führt es gegen Ihre Datenbank aus und liefert die Ergebnisse in Sekunden.",
         "zh": "用户能够将简单文本问题转化为完整的SQL查询并输出结果。只需输入您的问题（例如，展示上个季度前十名按收入排序的产品），SQL助理就会生成精确的SQL语句，对其运行您的数据库，并几秒钟内返回结果。"},
     "canvas_type": "Marketing",
     "dsl": {

agent/templates/trip_planner.json

@@ -2,13 +2,13 @@
 {
     "id": 14,
     "title": {
-		"en": "Trip Planner",
+		"en": "Trip planner",
 		"de": "Reiseplaner",
-		"zh": "旅行规划"},
+		"zh": "旅行规划师"},
     "description": {
 		"en": "This smart trip planner utilizes LLM technology to automatically generate customized travel itineraries, with optional tool integration for enhanced reliability.",
 		"de": "Dieser intelligente Reiseplaner nutzt LLM-Technologie zur automatischen Generierung maßgeschneiderter Reiserouten mit optionaler Tool-Integration für erhöhte Zuverlässigkeit.",
-		"zh": "智能旅行规划将利用大模型自动生成定制化的旅行行程，附带可选工具集成，以增强可靠性。"},
+		"zh": "智能旅行规划师将利用大模型自动生成定制化的旅行行程，附带可选工具集成，以增强可靠性。"},
 	"canvas_type": "Consumer App",
     "dsl": {
 		"components": {

agent/templates/your_starter_dataset_chatbot.json

@@ -2,14 +2,14 @@
 {
     "id": 9,
     "title": {
-        "en": "Technical Docs QA",
-        "de": "Technische Dokumentation Fragen & Antworten",
-        "zh": "技术文档问答"},
+        "en": "Your starter dataset chatbot",
+        "de": "Dein Starter-Datensatz-Chatbot",
+        "zh": "入门级知识库聊天助手"},
     "description": {
         "en": "This is a document question-and-answer system based on a knowledge base. When a user asks a question, it retrieves relevant document content to provide accurate answers.",
         "de": "Dies ist ein dokumentenbasiertes Frage-und-Antwort-System auf Basis einer Wissensdatenbank. Wenn ein Benutzer eine Frage stellt, werden relevante Dokumenteninhalte abgerufen, um genaue Antworten zu liefern.",
-        "zh": "基于知识库的文档问答系统，当用户提出问题时，会检索相关本地文档并提供准确回答。"},
-    "canvas_type": "Customer Support",
+        "zh": "基于知识库的入门级知识库聊天助手，当用户提出问题时，会检索相关本地文档并提供准确回答。"},
+    "canvas_type": "Recommended",
     "dsl": {
             "components": {
                 "Agent:StalePandasDream": {

agent/tools/base.py

@@ -57,17 +57,19 @@ class LLMToolPluginCallSession(ToolCallSession):
 
     async def tool_call_async(self, name: str, arguments: dict[str, Any]) -> Any:
         assert name in self.tools_map, f"LLM tool {name} does not exist"
+        logging.info(f"[ToolCall] invoke name={name} arguments={str(arguments)[:200]}")
         st = timer()
         tool_obj = self.tools_map[name]
         if isinstance(tool_obj, MCPToolCallSession):
             resp = await thread_pool_exec(tool_obj.tool_call, name, arguments, 60)
+        elif hasattr(tool_obj, "invoke_async") and asyncio.iscoroutinefunction(tool_obj.invoke_async):
+            resp = await tool_obj.invoke_async(**arguments)
         else:
-            if hasattr(tool_obj, "invoke_async") and asyncio.iscoroutinefunction(tool_obj.invoke_async):
-                resp = await tool_obj.invoke_async(**arguments)
-            else:
-                resp = await thread_pool_exec(tool_obj.invoke, **arguments)
+            resp = await thread_pool_exec(tool_obj.invoke, **arguments)
 
-        self.callback(name, arguments, resp, elapsed_time=timer()-st)
+        elapsed = timer() - st
+        logging.info(f"[ToolCall] done name={name} elapsed={elapsed:.2f}s result={str(resp)[:200]}")
+        self.callback(name, arguments, resp, elapsed_time=elapsed)
         return resp
 
     def get_tool_obj(self, name):
@@ -101,13 +103,8 @@ class ToolParamBase(ComponentParamBase):
             if "enum" in p:
                 params[k]["enum"] = p["enum"]
 
-        desc = self.meta["description"]
-        if hasattr(self, "description"):
-            desc = self.description
-
-        function_name = self.meta["name"]
-        if hasattr(self, "function_name"):
-            function_name = self.function_name
+        desc = getattr(self, "description", None) or self.meta["description"]
+        function_name = getattr(self, "function_name", self.meta["name"])
 
         return {
             "type": "function",

agent/tools/code_exec.py

@@ -18,15 +18,196 @@ import base64
 import json
 import logging
 import os
+import uuid
 from abc import ABC
+from collections.abc import Mapping
 from typing import Optional
 
 from pydantic import BaseModel, Field, field_validator
 from strenum import StrEnum
 
 from agent.tools.base import ToolBase, ToolMeta, ToolParamBase
+from api.db.services.file_service import FileService
 from common import settings
 from common.connection_utils import timeout
+from common.constants import SANDBOX_ARTIFACT_BUCKET, SANDBOX_ARTIFACT_EXPIRE_DAYS
+
+
+SYSTEM_OUTPUT_KEYS = frozenset(
+    {
+        "content",
+        "actual_type",
+        "_ERROR",
+        "_ARTIFACTS",
+        "_ATTACHMENT_CONTENT",
+        "raw_result",
+        "_created_time",
+        "_elapsed_time",
+    }
+)
+
+
+class ContractError(ValueError):
+    pass
+
+
+def _validate_business_output_name(name: str) -> None:
+    if not name or not name.strip():
+        raise ContractError("CodeExec business output name must not be empty")
+    if name in SYSTEM_OUTPUT_KEYS:
+        raise ContractError(f"CodeExec reserved output name is not allowed: {name}")
+    if "." in name:
+        raise ContractError(f"CodeExec business output name must not contain '.': {name}")
+
+
+def select_business_output(outputs: Mapping[str, object]) -> tuple[str, object]:
+    if len(outputs) == 1:
+        only_name, only_meta = next(iter(outputs.items()))
+        _validate_business_output_name(only_name)
+        return only_name, only_meta
+
+    business_outputs = [(name, meta) for name, meta in outputs.items() if name not in SYSTEM_OUTPUT_KEYS]
+    if len(business_outputs) != 1:
+        raise ContractError(
+            f"CodeExec contract must contain exactly one business output, got {len(business_outputs)}"
+        )
+    _validate_business_output_name(business_outputs[0][0])
+    return business_outputs[0]
+
+
+def normalize_output_value(value):
+    if isinstance(value, (tuple, list)):
+        return [normalize_output_value(item) for item in value]
+    if isinstance(value, dict):
+        return {key: normalize_output_value(item) for key, item in value.items()}
+    return value
+
+
+def infer_actual_type(value) -> str:
+    value = normalize_output_value(value)
+    if value is None:
+        return "Null"
+    if isinstance(value, bool):
+        return "Boolean"
+    if _is_number(value):
+        return "Number"
+    if isinstance(value, str):
+        return "String"
+    if isinstance(value, dict):
+        return "Object"
+    if isinstance(value, list):
+        if not value:
+            return "Array<Any>"
+        inferred = {infer_actual_type(item) for item in value}
+        if len(inferred) == 1:
+            return f"Array<{inferred.pop()}>"
+        return "Array<Any>"
+    return "Any"
+
+
+def render_canonical_content(value) -> str:
+    value = normalize_output_value(value)
+    if value is None:
+        return ""
+    if isinstance(value, str):
+        return value
+    if isinstance(value, (dict, list)):
+        return json.dumps(value, ensure_ascii=False, indent=2, sort_keys=True)
+    return str(value)
+
+
+def _is_number(value) -> bool:
+    return isinstance(value, (int, float)) and not isinstance(value, bool)
+
+
+def _validate_top_level_value_domain(value) -> None:
+    allowed = value is None or isinstance(value, (bool, str, dict, list)) or _is_number(value)
+    if not allowed:
+        raise ContractError(
+            f"CodeExec unsupported top-level result type: {type(value).__name__}. "
+            "Allowed top-level values are String, Number, Boolean, Object, Array, or Null."
+        )
+
+
+def _normalize_expected_type(expected_type: str) -> str:
+    etype = expected_type.strip()
+    low = etype.lower()
+    simple_types = {
+        "string": "String",
+        "number": "Number",
+        "boolean": "Boolean",
+        "object": "Object",
+        "null": "Null",
+        "any": "Any",
+    }
+    if low in simple_types:
+        return simple_types[low]
+    if low.startswith("array<") and low.endswith(">"):
+        inner = etype[etype.find("<") + 1 : -1].strip()
+        if not inner:
+            raise ContractError(f"Unsupported expected type: {expected_type}")
+        return f"Array<{_normalize_expected_type(inner)}>"
+    return etype
+
+
+def _validate_expected_type(expected_type: str, value, path: str = "") -> None:
+    etype = _normalize_expected_type(expected_type)
+    if not etype or etype.lower() == "any":
+        return
+
+    value = normalize_output_value(value)
+
+    if etype.startswith("Array<") and etype.endswith(">"):
+        inner_type = etype[6:-1].strip()
+        if not isinstance(value, list):
+            raise ContractError(
+                f"CodeExec contract mismatch at {path or 'value'}: expected type {etype}, got {infer_actual_type(value)}"
+            )
+        for index, item in enumerate(value):
+            child_path = f"{path}[{index}]" if path else f"[{index}]"
+            _validate_expected_type(inner_type, item, child_path)
+        return
+
+    actual_type = infer_actual_type(value)
+    if etype == "String":
+        valid = isinstance(value, str)
+    elif etype == "Number":
+        valid = _is_number(value)
+    elif etype == "Boolean":
+        valid = isinstance(value, bool)
+    elif etype == "Object":
+        valid = isinstance(value, dict)
+    elif etype == "Null":
+        valid = value is None
+    else:
+        raise ContractError(f"Unsupported expected type: {expected_type}")
+
+    if not valid:
+        raise ContractError(
+            f"CodeExec contract mismatch at {path or 'value'}: expected type {etype}, got {actual_type}"
+        )
+
+
+def build_code_exec_contract(outputs: Mapping[str, object], raw_result) -> dict[str, object]:
+    business_name, business_meta = select_business_output(outputs)
+    expected_type = ""
+    if isinstance(business_meta, Mapping):
+        expected_type = str(business_meta.get("type") or "")
+
+    normalized_value = normalize_output_value(raw_result)
+    _validate_top_level_value_domain(normalized_value)
+    _validate_expected_type(expected_type, normalized_value)
+
+    return {
+        "business_output": business_name,
+        "value": normalized_value,
+        "actual_type": infer_actual_type(normalized_value),
+        "content": render_canonical_content(normalized_value),
+    }
+
+
+def _art_field(art, field: str, default=""):
+    return art.get(field, default) if isinstance(art, dict) else getattr(art, field, default)
 
 
 class Language(StrEnum):
@@ -70,6 +251,7 @@ class CodeExecParam(ToolParamBase):
             "name": "execute_code",
             "description": """
 This tool has a sandbox that can execute code written in 'Python'/'Javascript'. It receives a piece of code and return a Json string.
+
 Here's a code example for Python(`main` function MUST be included):
 def main() -> dict:
     \"\"\"
@@ -84,6 +266,26 @@ def main() -> dict:
         "result": fibonacci_recursive(100),
     }
 
+To generate charts or files (images, PDFs, CSVs, etc.), save them to the `artifacts/` directory (relative to the working directory). The sandbox will automatically collect these files and return them. Example:
+def main() -> dict:
+    import matplotlib
+    matplotlib.use("Agg")
+    import matplotlib.pyplot as plt
+    import pandas as pd
+
+    df = pd.DataFrame({"x": [1, 2, 3, 4], "y": [10, 20, 25, 30]})
+    fig, ax = plt.subplots()
+    ax.plot(df["x"], df["y"])
+    ax.set_title("Sample Chart")
+    fig.savefig("artifacts/chart.png", dpi=150, bbox_inches="tight")
+    plt.close(fig)
+    return {"summary": "Chart saved to artifacts/chart.png"}
+
+Available Python packages: pandas, numpy, matplotlib, requests.
+Supported artifact file types: .png, .jpg, .jpeg, .svg, .pdf, .csv, .json, .html
+
+Collected artifacts are also parsed automatically and appended to the stable text output `content`. The content includes sections like `attachment1 (image): ...`, `attachment2 (pdf): ...`, so downstream nodes can consume a single text output without depending on unstable attachment-specific variables.
+
 Here's a code example for Javascript(`main` function MUST be included and exported):
 const axios = require('axios');
 async function main(args) {
@@ -125,6 +327,7 @@ module.exports = { main };
 
 class CodeExec(ToolBase, ABC):
     component_name = "CodeExec"
+    _lifecycle_configured = False
 
     @timeout(int(os.environ.get("COMPONENT_EXEC_TIMEOUT", 10 * 60)))
     def _invoke(self, **kwargs):
@@ -148,6 +351,8 @@ class CodeExec(ToolBase, ABC):
         if self.check_if_canceled("CodeExec execution"):
             return self.output()
 
+        timeout_seconds = int(os.environ.get("COMPONENT_EXEC_TIMEOUT", 10 * 60))
+
         try:
             # Try using the new sandbox provider system first
             try:
@@ -157,25 +362,19 @@ class CodeExec(ToolBase, ABC):
                     return
 
                 # Execute code using the provider system
-                result = sandbox_execute_code(
-                    code=code,
-                    language=language,
-                    timeout=int(os.environ.get("COMPONENT_EXEC_TIMEOUT", 10 * 60)),
-                    arguments=arguments
-                )
+                result = sandbox_execute_code(code=code, language=language, timeout=timeout_seconds, arguments=arguments)
 
                 if self.check_if_canceled("CodeExec execution"):
                     return
 
-                # Process the result
-                if result.stderr:
-                    self.set_output("_ERROR", result.stderr)
-                    return
-
-                parsed_stdout = self._deserialize_stdout(result.stdout)
-                logging.info(f"[CodeExec]: Provider system -> {parsed_stdout}")
-                self._populate_outputs(parsed_stdout, result.stdout)
-                return
+                artifacts = result.metadata.get("artifacts", []) if result.metadata else []
+                return self._process_execution_result(
+                    result.stdout,
+                    result.stderr,
+                    "Provider system",
+                    artifacts,
+                    execution_metadata=result.metadata,
+                )
 
             except (ImportError, RuntimeError) as provider_error:
                 # Provider system not available or not configured, fall back to HTTP
@@ -196,7 +395,7 @@ class CodeExec(ToolBase, ABC):
                 self.set_output("_ERROR", "Task has been canceled")
                 return self.output()
 
-            resp = requests.post(url=f"http://{settings.SANDBOX_HOST}:9385/run", json=code_req, timeout=int(os.environ.get("COMPONENT_EXEC_TIMEOUT", 10 * 60)))
+            resp = requests.post(url=f"http://{settings.SANDBOX_HOST}:9385/run", json=code_req, timeout=timeout_seconds)
             logging.info(f"http://{settings.SANDBOX_HOST}:9385/run,  code_req: {code_req}, resp.status_code {resp.status_code}:")
 
             if self.check_if_canceled("CodeExec execution"):
@@ -206,14 +405,13 @@ class CodeExec(ToolBase, ABC):
                 resp.raise_for_status()
             body = resp.json()
             if body:
-                stderr = body.get("stderr")
-                if stderr:
-                    self.set_output("_ERROR", stderr)
-                    return self.output()
-                raw_stdout = body.get("stdout", "")
-                parsed_stdout = self._deserialize_stdout(raw_stdout)
-                logging.info(f"[CodeExec]: http://{settings.SANDBOX_HOST}:9385/run -> {parsed_stdout}")
-                self._populate_outputs(parsed_stdout, raw_stdout)
+                return self._process_execution_result(
+                    body.get("stdout", ""),
+                    body.get("stderr"),
+                    f"http://{settings.SANDBOX_HOST}:9385/run",
+                    body.get("artifacts", []),
+                    execution_metadata=self._build_http_execution_metadata(body),
+                )
             else:
                 self.set_output("_ERROR", "There is no response from sandbox")
                 return self.output()
@@ -226,6 +424,129 @@ class CodeExec(ToolBase, ABC):
 
         return self.output()
 
+    def _process_execution_result(
+        self,
+        stdout: str,
+        stderr: str | None,
+        source: str,
+        artifacts: list | None = None,
+        execution_metadata: dict | None = None,
+    ):
+        has_structured_result = bool((execution_metadata or {}).get("result_present") is True)
+        resolved_value, used_stdout_fallback = self._resolve_execution_result_value(stdout, execution_metadata)
+
+        if stderr and not has_structured_result and not artifacts and not str(stdout or "").strip():
+            self.set_output("_ERROR", stderr)
+            return self.output()
+
+        # Clear any stale error from previous runs or base class initialization
+        self.set_output("_ERROR", "")
+
+        if stderr:
+            logging.warning(f"[CodeExec]: stderr (non-fatal): {stderr[:500]}")
+
+        if used_stdout_fallback and str(stdout or "").strip():
+            logging.warning("[CodeExec]: Falling back to stdout deserialization because no structured result metadata was provided")
+
+        logging.info(f"[CodeExec]: {source} -> {resolved_value}")
+        content_parts = []
+        base_content = self._apply_business_output(resolved_value)
+        if base_content:
+            content_parts.append(base_content)
+
+        if artifacts:
+            artifact_urls = self._upload_artifacts(artifacts)
+            self.set_output("_ARTIFACTS", artifact_urls or None)
+            attachment_text = self._build_attachment_content(artifacts, artifact_urls)
+            self.set_output("_ATTACHMENT_CONTENT", attachment_text)
+            if attachment_text:
+                content_parts.append(attachment_text)
+        else:
+            self.set_output("_ARTIFACTS", None)
+            self.set_output("_ATTACHMENT_CONTENT", "")
+
+        self.set_output("content", "\n\n".join([part for part in content_parts if part]).strip())
+
+        return self.output()
+
+    def _build_http_execution_metadata(self, body: Mapping | None) -> dict:
+        if not isinstance(body, Mapping):
+            return {}
+        structured_result = body.get("result")
+        if not isinstance(structured_result, Mapping):
+            return {}
+        return {
+            "result_present": structured_result.get("present", False),
+            "result_value": structured_result.get("value"),
+            "result_type": structured_result.get("type"),
+        }
+
+    def _resolve_execution_result_value(self, stdout: str, execution_metadata: Mapping | None = None):
+        metadata = execution_metadata or {}
+        if metadata.get("result_present") is True:
+            return metadata.get("result_value"), False
+        return self._deserialize_stdout(stdout), True
+
+    @classmethod
+    def _ensure_bucket_lifecycle(cls):
+        if cls._lifecycle_configured:
+            return
+        try:
+            storage = settings.STORAGE_IMPL
+            # Only MinIO/S3 backends expose .conn for lifecycle config
+            if not hasattr(storage, "conn") or storage.conn is None:
+                cls._lifecycle_configured = True
+                return
+            if not storage.conn.bucket_exists(SANDBOX_ARTIFACT_BUCKET):
+                storage.conn.make_bucket(SANDBOX_ARTIFACT_BUCKET)
+            from minio.commonconfig import Filter
+            from minio.lifecycleconfig import Expiration, LifecycleConfig, Rule
+
+            rule = Rule(
+                rule_id="auto-expire",
+                status="Enabled",
+                rule_filter=Filter(prefix=""),
+                expiration=Expiration(days=SANDBOX_ARTIFACT_EXPIRE_DAYS),
+            )
+            storage.conn.set_bucket_lifecycle(SANDBOX_ARTIFACT_BUCKET, LifecycleConfig([rule]))
+            logging.info(f"[CodeExec]: Set {SANDBOX_ARTIFACT_EXPIRE_DAYS}-day lifecycle on bucket '{SANDBOX_ARTIFACT_BUCKET}'")
+            cls._lifecycle_configured = True
+        except Exception as e:
+            # Do NOT set _lifecycle_configured so we retry next time
+            logging.warning(f"[CodeExec]: Failed to set bucket lifecycle: {e}")
+
+    def _upload_artifacts(self, artifacts: list) -> list[dict]:
+        self._ensure_bucket_lifecycle()
+        uploaded = []
+        for art in artifacts:
+            try:
+                name = _art_field(art, "name")
+                content_b64 = _art_field(art, "content_b64")
+                mime_type = _art_field(art, "mime_type")
+                size = _art_field(art, "size", 0)
+                if not content_b64 or not name:
+                    continue
+
+                ext = os.path.splitext(name)[1].lower()
+                storage_name = f"{uuid.uuid4().hex}{ext}"
+                binary = base64.b64decode(content_b64)
+
+                settings.STORAGE_IMPL.put(SANDBOX_ARTIFACT_BUCKET, storage_name, binary)
+
+                url = f"/v1/document/artifact/{storage_name}"
+                uploaded.append(
+                    {
+                        "name": name,
+                        "url": url,
+                        "mime_type": mime_type,
+                        "size": size,
+                    }
+                )
+                logging.info(f"[CodeExec]: Uploaded artifact {name} -> {url}")
+            except Exception as e:
+                logging.warning(f"[CodeExec]: Failed to upload artifact: {e}")
+        return uploaded
+
     def _encode_code(self, code: str) -> str:
         return base64.b64encode(code.encode("utf-8")).decode("utf-8")
 
@@ -243,139 +564,84 @@ class CodeExec(ToolBase, ABC):
                 continue
         return text
 
-    def _coerce_output_value(self, value, expected_type: Optional[str]):
-        if expected_type is None:
-            return value
-
-        etype = expected_type.strip().lower()
-        inner_type = None
-        if etype.startswith("array<") and etype.endswith(">"):
-            inner_type = etype[6:-1].strip()
-            etype = "array"
+    def _apply_business_output(self, parsed_stdout) -> str:
+        normalized_result = normalize_output_value(parsed_stdout)
+        self.set_output("raw_result", normalized_result)
 
+        business_output_names = [name for name in self._param.outputs if name not in SYSTEM_OUTPUT_KEYS]
         try:
-            if etype == "string":
-                return "" if value is None else str(value)
+            contract = build_code_exec_contract(self._param.outputs, normalized_result)
+        except ContractError as e:
+            for output_name in business_output_names:
+                self.set_output(output_name, None)
+            self.set_output("actual_type", infer_actual_type(normalized_result))
+            self.set_output("_ERROR", str(e))
+            logging.warning(f"[CodeExec]: contract validation failed: {e}")
+            return render_canonical_content(normalized_result)
 
-            if etype == "number":
-                if value is None or value == "":
-                    return None
-                if isinstance(value, (int, float)):
-                    return value
-                if isinstance(value, str):
-                    try:
-                        return float(value)
-                    except Exception:
-                        return value
-                return float(value)
+        self.set_output("actual_type", contract["actual_type"])
+        self.set_output(contract["business_output"], contract["value"])
+        return contract["content"]
 
-            if etype == "boolean":
-                if isinstance(value, bool):
-                    return value
-                if isinstance(value, str):
-                    lv = value.lower()
-                    if lv in ("true", "1", "yes", "y", "on"):
-                        return True
-                    if lv in ("false", "0", "no", "n", "off"):
-                        return False
-                return bool(value)
+    def _build_attachment_content(self, artifacts: list, artifact_urls: list[dict] | None = None) -> str:
+        sections = []
+        artifact_urls = artifact_urls or []
 
-            if etype == "array":
-                candidate = value
-                if isinstance(candidate, str):
-                    parsed = self._deserialize_stdout(candidate)
-                    candidate = parsed
-                if isinstance(candidate, tuple):
-                    candidate = list(candidate)
-                if not isinstance(candidate, list):
-                    candidate = [] if candidate is None else [candidate]
-
-                if inner_type == "string":
-                    return ["" if v is None else str(v) for v in candidate]
-                if inner_type == "number":
-                    coerced = []
-                    for v in candidate:
-                        try:
-                            if v is None or v == "":
-                                coerced.append(None)
-                            elif isinstance(v, (int, float)):
-                                coerced.append(v)
-                            else:
-                                coerced.append(float(v))
-                        except Exception:
-                            coerced.append(v)
-                    return coerced
-                return candidate
-
-            if etype == "object":
-                if isinstance(value, dict):
-                    return value
-                if isinstance(value, str):
-                    parsed = self._deserialize_stdout(value)
-                    if isinstance(parsed, dict):
-                        return parsed
-                return value
-        except Exception:
-            return value
-
-        return value
-
-    def _populate_outputs(self, parsed_stdout, raw_stdout: str):
-        outputs_items = list(self._param.outputs.items())
-        logging.info(f"[CodeExec]: outputs schema keys: {[k for k, _ in outputs_items]}")
-        if not outputs_items:
-            return
-
-        if isinstance(parsed_stdout, dict):
-            for key, meta in outputs_items:
-                if key.startswith("_"):
+        for idx, art in enumerate(artifacts, start=1):
+            key = f"attachment{idx}"
+            try:
+                name = _art_field(art, "name")
+                content_b64 = _art_field(art, "content_b64")
+                mime_type = _art_field(art, "mime_type")
+                if not name or not content_b64:
                     continue
-                val = self._get_by_path(parsed_stdout, key)
-                if val is None and len(outputs_items) == 1:
-                    val = parsed_stdout
-                coerced = self._coerce_output_value(val, meta.get("type"))
-                logging.info(f"[CodeExec]: populate dict key='{key}' raw='{val}' coerced='{coerced}'")
-                self.set_output(key, coerced)
-            return
 
-        if isinstance(parsed_stdout, (list, tuple)):
-            for idx, (key, meta) in enumerate(outputs_items):
-                if key.startswith("_"):
-                    continue
-                val = parsed_stdout[idx] if idx < len(parsed_stdout) else None
-                coerced = self._coerce_output_value(val, meta.get("type"))
-                logging.info(f"[CodeExec]: populate list key='{key}' raw='{val}' coerced='{coerced}'")
-                self.set_output(key, coerced)
-            return
+                blob = base64.b64decode(content_b64)
+                parsed = FileService.parse(
+                    name,
+                    blob,
+                    False,
+                    tenant_id=self._canvas.get_tenant_id(),
+                )
+                attachment_type = self._normalize_attachment_type(name, mime_type)
+                section = self._format_attachment_section(key, attachment_type, name, parsed)
+                sections.append(section)
+                logging.info(f"[CodeExec]: parse attachment section key='{key}' from artifact='{name}'")
+            except Exception as e:
+                logging.warning(f"[CodeExec]: Failed to parse artifact for content section '{key}': {e}")
+                fallback_type = self._normalize_attachment_type(name, mime_type)
+                fallback_name = name
+                fallback_url = ""
+                if idx - 1 < len(artifact_urls):
+                    fallback_url = artifact_urls[idx - 1].get("url", "")
+                fallback_text = "Artifact generated but parse failed."
+                if fallback_url:
+                    fallback_text += f" Download: {fallback_url}"
+                sections.append(self._format_attachment_section(key, fallback_type, fallback_name, fallback_text))
 
-        default_val = parsed_stdout if parsed_stdout is not None else raw_stdout
-        for idx, (key, meta) in enumerate(outputs_items):
-            if key.startswith("_"):
-                continue
-            val = default_val if idx == 0 else None
-            coerced = self._coerce_output_value(val, meta.get("type"))
-            logging.info(f"[CodeExec]: populate scalar key='{key}' raw='{val}' coerced='{coerced}'")
-            self.set_output(key, coerced)
+        if sections:
+            return f"attachment_count: {len(sections)}\n\n" + "\n\n".join(sections)
+        return "attachment_count: 0"
 
-    def _get_by_path(self, data, path: str):
-        if not path:
-            return None
-        cur = data
-        for part in path.split("."):
-            part = part.strip()
-            if not part:
-                return None
-            if isinstance(cur, dict):
-                cur = cur.get(part)
-            elif isinstance(cur, list):
-                try:
-                    idx = int(part)
-                    cur = cur[idx]
-                except Exception:
-                    return None
-            else:
-                return None
-            if cur is None:
-                return None
-        logging.info(f"[CodeExec]: resolve path '{path}' -> {cur}")
-        return cur
+    def _normalize_attachment_type(self, name: str, mime_type: str) -> str:
+        mime_type = str(mime_type or "").strip().lower()
+        if mime_type.startswith("image/"):
+            return "image"
+        if mime_type == "application/pdf":
+            return "pdf"
+        if mime_type == "text/csv":
+            return "csv"
+        if mime_type == "application/json":
+            return "json"
+        if mime_type == "text/html":
+            return "html"
+
+        ext = os.path.splitext(name or "")[1].lower().lstrip(".")
+        return ext or "file"
+
+    def _format_attachment_section(self, key: str, attachment_type: str, name: str, parsed: str) -> str:
+        title = f"{key} ({attachment_type})"
+        if name:
+            title += f": {name}"
+        body = parsed if isinstance(parsed, str) else json.dumps(parsed, ensure_ascii=False)
+        return f"{title}\n{body}".strip()