## Summary
- Fix TOOL_CALL_RESULT events reusing parent assistant message ID across
StreamingToolAgent, SharedStateReadWriteController, and
SubagentsController —
React's deduplicateMessages() Map overwrote assistant messages with tool
messages when they shared the same ID
- Add `data-testid="copilot-assistant-message"` to BYOC hashbrown
renderer so
the D5 test harness can detect assistant messages
- Register AG-UI Jackson mixins via JacksonConfig postConfigurer,
normalize
array-format content via ContentNormalizingModule, remove ChatMemory
beans
that interfered with aimock fixture matching, inject Spring-managed
ObjectMapper into AgentConfigController
## Test plan
- [x] All 28/28 spring-ai D5 features pass locally (verified stable
across
multiple consecutive runs)
- [x] Docker build succeeds
- [x] No regressions — previously passing 22 features remain green
Root cause: TOOL_CALL_RESULT events reused the parent assistant
message ID, causing React deduplicateMessages() to overwrite
assistant messages with tool messages (Map keyed by ID). Fix uses
unique UUIDs for each tool result message across all three
controllers. Also adds data-testid to the BYOC hashbrown renderer
so the D5 harness can detect assistant messages, registers AG-UI
Jackson mixins via JacksonConfig postConfigurer, normalizes
array-format content via ContentNormalizingModule, removes ChatMemory
beans that interfered with aimock fixture matching, and injects
Spring-managed ObjectMapper into AgentConfigController.
Covers: failure triage (text-only vs tool features, probe text extraction,
message disappearing, chatMemory pollution), framework-specific gotchas
(LlamaIndex, Spring AI, Built-in Agent, MS Agent, AG2), and production
vs local parity checklist.
## Summary
Replaces the `/eval` PR comment trigger with a professional GitHub Check
Run button UX, and adds native (no-Docker) CI execution infrastructure.
### Check Run Button (Phase 1)
- **showcase_eval_check.yml** — creates a Check Run with "Run Showcase
Eval" action button on every PR open/push
- **showcase/eval-webhook/** — tiny Hono relay service on Railway that
receives `check_run.requested_action` webhooks, authenticates as devops
bot, and dispatches `showcase_eval.yml` via `workflow_dispatch`
- **showcase_eval.yml** — adds `workflow_dispatch` trigger with
`dispatch-gate` job, Check Run lifecycle updates (neutral → in_progress
→ completed with results), devops bot token for Checks API
### Native Execution (Phase 2)
- **`--ci` flag** on eval orchestrator — skips Docker lifecycle
(`up/down/isRunning/docker-inspect`), assumes services already running
- **ci-native-eval.sh** — standalone helper that installs deps, starts
`next dev --turbopack` + Python agents natively, health-waits, runs
`showcase eval --ci`
- **test_e2e-showcase-on-demand.yml** — fixed and extended with
`langgraph-python` support, agent-type detection (uvicorn vs
langgraph_cli), relaxed aimock_toggle.py requirement
### Setup required (post-merge)
1. Add `checks:write` permission to the copilotkit-devops-bot GitHub App
2. Deploy eval-webhook to Railway with app secrets
3. Configure GitHub App webhook URL to point to the Railway service
## Test plan
- [x] showcase-harness: 1480 tests pass
- [ ] CI green
- [ ] Deploy eval-webhook, verify `/health` endpoint
- [ ] Open test PR, verify Check Run appears with button
- [ ] Click button, verify eval triggers and Check Run updates
## Summary
- Adds Deep Agents as a top-level integration in the docs (selector,
grid, quickstart dropdown, feature matrix)
- Creates `docs/integrations/deepagents/` with landing page and
quickstart (content ported from `langgraph/deep-agents`)
- Adds `deepagents` rewrite rule in `next.config.mjs` so `/deepagents/*`
routes correctly
- Adds callout on `langgraph/deep-agents` page pointing to the new
integration for discoverability
Closes CPK-7187
## Test plan
- [ ] `/deepagents` loads the landing page
- [ ] `/deepagents/quickstart` loads the quickstart guide
- [ ] Deep Agents appears in the integration selector dropdown
- [ ] Deep Agents appears in the integration grid
- [ ] Deep Agents appears in the quickstart dropdown
- [ ] `/langgraph/deep-agents` shows the callout linking to
`/deepagents`
🤖 Generated with [Claude Code](https://claude.com/claude-code)
## Summary
Bumps `@ag-ui/client`, `@ag-ui/core`, and `@ag-ui/encoder` from `0.0.52`
→ `0.0.53` to pick up the ESM interop fix in
[ag-ui-protocol/ag-ui#1578](https://github.com/ag-ui-protocol/ag-ui/pull/1578).
## Why
`@ag-ui/client@0.0.52`'s ESM bundle compiled `import jsonpatch from
"fast-json-patch"` down to `import * as a from "fast-json-patch"`. Under
Node's native ESM loader (and webpack/turbopack in Next.js), that
namespace import is empty for `fast-json-patch@3.x` because the module
populates its exports via `Object.assign(exports, ...)` — the CJS→ESM
named-export detector cannot see them.
Result: `a.applyPatch is not a function` thrown on every `STATE_DELTA`
and `ACTIVITY_DELTA` event. The most visible symptom is open generative
UI with LangGraph agents, where the console floods with:
```
[ui] Failed to apply activity patch for 'call_xxx-activity': a.applyPatch is not a function
```
`0.0.53` switches the source back to a default import so the emitted
bundle works under both ESM and CJS consumers.
## Test plan
- [x] `pnpm install` resolves cleanly to
`@ag-ui/{client,core,encoder}@0.0.53` in the lockfile
- [x] Pre-commit hooks pass (lint, tests, package checks)
- [ ] Verify open generative UI with a LangGraph agent no longer logs
`applyPatch is not a function`
🤖 Generated with [Claude Code](https://claude.com/claude-code)
- Add buffer length check before timingSafeEqual to prevent RangeError
on missing/malformed X-Hub-Signature-256 headers
- Switch Dockerfile from pnpm to npm with package-lock.json (pnpm
lockfile lives at monorepo root, not in the package directory)
<!--
Thank you for sending the PR! We appreciate you spending the time to
work on these changes.
Help us understand your motivation by explaining why you decided to make
this change.
**Please PLEASE reach out to us first before starting any significant
work on new or existing features.**
By the time you've gotten here, you're looking at creating a pull
request so hopefully we're not too late.
We love community contributions! That said, we want to make sure we're
all on the same page before you start.
Investing a lot of time and effort just to find out it doesn't align
with the upstream project feels awful, and we don't want that to happen.
It also helps to make sure the work you're planning isn't already in
progress.
As described in our contributing guide, please file an issue first:
https://github.com/ag-ui-protocol/ag-ui/issues
Or, reach out to us on Discord: https://discord.com/invite/6dffbvGU3D
You can learn more about contributing to copilotkit here:
https://github.com/copilotkit/copilotkit/blob/master/CONTRIBUTING.md
Happy contributing!
-->
## What does this PR do?
(Describe the changes introduced in this PR)
## Related PRs and Issues
- (Direct link to related PR or issue, if relevant)
## Checklist
- [ ] I have read the [Contribution
Guide](https://github.com/copilotkit/copilotkit/blob/master/CONTRIBUTING.md)
- [ ] If the PR changes or adds functionality, I have updated the
relevant documentation
- [ ] "Allow edits by maintainers" is checked (lets us help iterate on
your PR directly — faster turnaround for everyone)
Add --ci flag to eval orchestrator that skips Docker lifecycle and
assumes services are already running. Add ci-native-eval.sh helper
that installs deps, starts next dev + agent servers natively, health-
waits, then runs showcase eval --ci. Fix on-demand E2E workflow with
langgraph-python support and agent-type detection.
New showcase_eval_check.yml creates a Check Run with "Run Showcase
Eval" action button on every PR. Modified showcase_eval.yml adds
workflow_dispatch trigger with dispatch-gate job, Check Run update
in post-result, and devops bot token for Checks API calls.
Hono web server that receives check_run.requested_action webhooks from
GitHub, authenticates as the devops bot, updates the Check Run to
in_progress, and dispatches showcase_eval.yml via workflow_dispatch.
Includes GHCR build workflow and pnpm workspace registration.
## Summary
- Add D2/API badge to per-cell status row in shell-dashboard (cells now
show `API ✓ RT ✓ CV ✓` instead of just `RT ✓ CV ✓`)
- Wire `d2` field in `CellState` to the existing `agent:<slug>` PB rows
from the agent-check probe
- Add API (Agent) dimension to the CellDrilldown popover panel
- Fix stale FP/D6 test references left over from #4513
## Test plan
- [x] `live-status.test.ts` — 3 new tests: d2
present/absent/rollup-exclusion (42 total, all pass)
- [x] `cell-drilldown.test.tsx` — updated "renders all 5 badge
dimensions" to check API (Agent) instead of FP
- [x] `overlay-selector-integration.test.tsx` — FP → API in health-only,
health+docs, testing-kind assertions
- [x] `cell-pieces.test.tsx` — CP8 tests updated from FP to API
- [x] `compute-tally-detail.test.tsx` — added `d2` to `makeCellState` to
satisfy TS
- [x] TypeScript type-check clean (no new errors)
The cell status row rendered RT (D4/e2e) and CV (D5) badges but was
missing the D2/API badge despite the legend already documenting it.
- Add `d2` field to `CellState` sourced from `agent:<slug>` rows
- Render API LiveBadge before RT in CellStatus (D2 < D4 ordering)
- Add API (Agent) to CellDrilldown DIMENSIONS array
- Fix stale FP/D6 references in tests left over from #4513
## Summary
`available: "always"` suggestion configs (static and dynamic) didn't
render on the welcome screen when the chat used `runtimeUrl` to fetch
agents instead of registering them locally with
`agents__unsafe_dev_only`.
The bug has been latent since v2 first landed (Dec 2025); a per-thread
cloning change masked it for some flows from Mar 31 → Apr 23, and the
Apr 23 revert (#3525 backout) re-exposed it.
This PR is welcome-screen only — the `!isConnecting && !isRunning` UI
gate is unchanged, so suggestions still hide during connect/replay and
during runs as before.
## What was broken
With `runtimeUrl`, the agents registry is empty during the initial
`/info` fetch. Two compounding issues meant `available: "always"`
configs never got off the ground:
1. **`SuggestionEngine.reloadSuggestions`** bailed early when the
consumer agent wasn't in the registry yet. The first reload fires from
`useConfigureSuggestions` on mount — at that moment the registry is
empty, so every config got skipped. Static pills never appeared, and
dynamic pills never even started generating.
2. **`useConfigureSuggestions`'s global-config path** (no
`consumerAgentId` or `"*"`) only iterated the current agents map. Empty
map → zero reload calls → suggestions stuck empty until something else
triggered a reload.
## Fix
Three small changes, scoped to the welcome screen path:
1. `SuggestionEngine.reloadSuggestions` no longer bails when the agent's
missing — defaults `messageCount` to 0 and processes static configs
anyway. Dynamic configs still skip until a real agent arrives.
2. `useConfigureSuggestions`'s global path also calls
`reloadSuggestions(targetAgentId)` directly (covers the empty-map case
where the agent the chat is bound to isn't yet in the registry).
3. `useConfigureSuggestions` subscribes to `onAgentsChanged` *only* for
dynamic configs, *only* when the target agent isn't yet present, and
*unsubscribes after firing once*. Dynamic pills catch up after the
runtime fetch completes, without piling up overlapping generations as
multiple hooks mount.
## What's preserved
- `hasSuggestions` keeps `!isConnecting && !isRunning` — bootstrap
replay and run-in-flight both still hide pills (no mid-replay layout
jump, no stale-context flash mid-run).
- `available: "always"` is the *eligibility window* (welcome screen vs
after first message), not a "render through transitions" override.
- Threading behavior (thread switch, explicit `threadId`, multi-chat) is
unchanged. None of the new code paths fire on thread connects.
## Test plan
- [x] `packages/core` — engine unit tests for `reloadSuggestions` when
no agent is present + when only static "always" config exists. All 59
core-suggestions tests pass.
- [x] `packages/react-core` — 4 integration tests in
`CopilotChat.suggestionsAlways.test.tsx`:
- shows on welcome screen with explicit `consumerAgentId`
- shows on welcome screen with global config
- hides during a run, reappears after (default scroll)
- hides during a run, reappears after (pin-to-send)
- [x] All 1157 react-core tests pass.
- [ ] Manual smoke in `examples/v2/react/demo` with both static and
dynamic `available: "always"` configs — welcome screen pills, hide
during run, regenerate after.
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Two-part welcome-screen regression for `available: "always"` suggestion configs
when the chat connects to agents via `runtimeUrl` instead of registering them
locally:
1. SuggestionEngine.reloadSuggestions bailed early when the agent wasn't yet
in the registry. With runtimeUrl, the registry is empty during the initial
/info fetch, so the very first reload (fired by useConfigureSuggestions on
mount) skipped every config — static pills never appeared on the welcome
screen, dynamic pills never started generating. Now: don't bail, default
`messageCount` to 0, run static configs anyway. Dynamic configs still need
a real agent and skip until one arrives.
2. useConfigureSuggestions's global-config path only iterated the current
agents map, which compounded the problem above — the empty map meant zero
reloads. Now: also reload for the chat's resolved consumer agent (covers
the empty-map case), and subscribe to onAgentsChanged for dynamic configs
only, firing exactly once when the target agent first appears (so dynamic
pills catch up after the runtime fetch completes, without piling up
overlapping generations as multiple useConfigureSuggestions hooks mount).
`hasSuggestions` keeps the `!isConnecting && !isRunning` UI gate. `available:
"always"` controls eligibility windows (welcome screen vs. after first
message), not whether to render through connect/replay or through a run —
those still hide and the end-of-run reload regenerates against the new
context.
Tests:
- Engine: added unit coverage for reloadSuggestions when no agent is present.
- React: 4 integration tests covering welcome screen (specific + global
consumerAgentId) and the run lifecycle (hide during run, reappear after) in
default and pin-to-send modes.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
## Summary
Critical-path content fixes from the shell-docs QA triage. Fixes
everything that breaks copy-paste or click-through on existing pages.
**Items addressed:**
- 2.2, 8.1, 13.1, 13.2 — code-block import hygiene
- 12.1, 14.1, 15.1, 15.2, 22.1 — cross-page link sweep
- 9.1, 9.2 — `gpt-5.2*` → `gpt-5.4*` sweep + extend model-name CI
validator
- 7.1 — `migrations.enabled` premium walkthrough corrections
## Visual inspection
1. Start the dev server:
```bash
nx run shell-docs:dev
```
Open http://localhost:3003.
2. **Imports — copy-paste check.** Visit each page below; copy each
`route.ts` / `app/layout.tsx` / `app/page.tsx` block into a scratch
TypeScript file and verify it has all needed imports:
- `/built-in-agent/quickstart` (BIA, default integration — should now
import `CopilotKit` from `@copilotkit/react-core/v2`)
- `/agent-spec/quickstart`
- `/microsoft-agent-framework/quickstart`
- Switch framework picker to each of: `langgraph`, `mastra`,
`pydantic-ai`, `adk`, `agno`, `aws-strands`, `llamaindex`. Re-check the
same blocks.
- `/auth`, `/agentic-protocols/mcp`, `/backend/copilot-runtime`,
`/multimodal-attachments` — verify code blocks now show import lines.
3. **Links — click-through check.** On each page below, click every
external link in the prose:
- `/agentic-protocols` (the index — 7 link rewrites; all should resolve)
- `/multimodal-attachments` (the migration Callout should be GONE)
- `/faq` ("What's available?" table — the rows should be plain bold
labels, not links)
- `/langgraph/agent-app-context` (mid-prose link should now point at
langgraph fw page, not BIA)
- `/inspector` (the agent-app-context relative link should resolve to
langgraph fw page)
- `/troubleshooting/common-issues` (the `copilot-runtime` and
`model-selection` links should now resolve)
4. **Model names — search check.** In the dev server (or via grep),
confirm `gpt-5.2` and `gpt-5.2-mini` no longer appear anywhere in
`showcase/shell-docs/src/content/`. All references should now be
`gpt-5.4` / `gpt-5.4-mini`.
5. **Premium walkthrough — read-through.** Walk `/premium/self-hosting`
end to end as if installing fresh. The walkthrough should now tell you
to set `migrations.enabled: true` BEFORE the install command. The "Job
will appear as Completed" prose should only fire if migrations were
enabled.
6. **Anti-checks (these should NOT have changed):**
- The legitimate `gpt-4o`, `gpt-4.1`, `gpt-5.4` references in JSDoc /
source code.
- Code blocks where imports were intentionally omitted because a prior
block on the same page established them.
- Reference pages under `/reference/v1/` (those are owned by a separate
PR).
## Summary
Sidebar / IA / HITL cleanup from the shell-docs QA triage. Adds fallback
callouts to HITL pages for non-native frameworks, fixes gate scoping
bugs, wires orphan pages into nav, adds missing `meta.json`.
**Items addressed:**
- 3.1 — `useInterrupt` and `headless` fallback callout for the 13
frameworks without `interrupt_pattern` (via new `absent` mode on
`<WhenFrameworkHas>`)
- 3.3 — `headless.mdx` `useHeadlessInterrupt` symbol scope fix (wrap
"Driving it from plain UI" inside the native gate)
- 3.5 — `multi-agent/meta.json` added so breadcrumbs use the explicit
"Multi-Agent" label
- 17.1 — `agent-config.mdx` shared lead-in between `<InlineDemo>` and
the gated branches
- 18.1 — `ag-ui-middleware.mdx` moved under `agentic-protocols/`,
registered in `meta.json`, links to upstream AG-UI guide, 302 from the
old `/ag-ui-middleware` path
## Visual inspection
1. Start the dev server:
```bash
nx run shell-docs:dev
```
Open http://localhost:3003.
2. **HITL fallback callout.** Switch the framework picker to each of:
`built-in-agent`, `mastra`, `crewai-crews`, `ag2`, `google-adk`,
`pydantic-ai`, `agno`, `claude-sdk-python`, `claude-sdk-typescript`,
`llamaindex`, `langroid`, `spring-ai`, `strands`. For each, navigate to:
- `/human-in-the-loop/useInterrupt`
- `/human-in-the-loop/headless`
The page should render a callout pointing readers at
`/human-in-the-loop` (`useHumanInTheLoop`), NOT collapse to an empty
middle.
3. **HITL native framework check (no regression).** Switch to
`langgraph-fastapi`, `langgraph-typescript`, `langgraph-python`,
`ms-agent-dotnet`, `ms-agent-python`. Same two pages should still render
their full native/promise-based content. Especially check
`/human-in-the-loop/headless` for proper `useHeadlessInterrupt` usage
now within its gate.
4. **Multi-agent sidebar.** With any framework selected, confirm
"Multi-Agent" appears in the sidebar with `subagents` as a sub-entry.
5. **agent-config lead-in.** Visit `/agent-config`. Verify there's a
smooth prose transition between the InlineDemo and the "How it works"
section.
6. **AG-UI middleware sidebar.** Confirm `ag-ui-middleware` appears in
the sidebar under the AG-UI / agentic-protocols section. Click through
and verify the page renders correctly. Verify the 302 from the old
`/ag-ui-middleware` URL works.
7. **Anti-checks (should NOT have changed):**
- Per-framework HITL nav for the 5 native fws unchanged in shape.
- Other sidebar sections unaffected.
- Existing `<WhenFrameworkHas>` gates on other pages still work.
## Summary
- Reorder dashboard legend so all depth/level explanations (D0-D4, L1-L4
Strip, D4 RT, D5 CV, D6 FP) are grouped together on the left
- Regression indicator, color chips, and status symbols shift right
- Moved `▼ depth regression` from DepthLegend into HealthLegend to keep
it after the FP explanation
## Test plan
- [ ] Visual check: legend items render in new order with depth
explanations grouped left
Reorder legend items so depth-related explanations (D0-D4, L1-L4 Strip,
D4 RT, D5 CV, D6 FP) are grouped together on the left, followed by the
regression indicator, color chips, and status symbols.
## Summary
- The DepthChip (colored badge) correctly shows D4/D5/D6 depth numbers
-- unchanged.
- The descriptive text row below each cell's chip was duplicating the
D4/D5/D6 labels instead of showing the human-readable RT (Round Trip) /
CV (Conversation) / FP (Feature Parity) abbreviations.
- Updates LiveBadge name props in CellStatus, DIMENSIONS labels in
CellDrilldown, and all related tests.
## Test plan
- [x] All overlay-selector-integration tests pass (15 tests)
- [x] All cell-drilldown tests pass (7 tests)
- [x] cell-pieces tests pass (pre-existing timestamp format failures
only, confirmed identical on main)
The DepthChip (colored badge) correctly shows D4/D5/D6 depth numbers.
The descriptive text row below each cell's chip was also showing D4/D5/D6
instead of the human-readable RT (Round Trip) / CV (Conversation) /
FP (Feature Parity) labels. Fix the LiveBadge name props in CellStatus,
the DIMENSIONS labels in CellDrilldown, and all related tests.
## Summary
Partially reverts #4506: badge chips now display depth-layer numbers
(D4, D5, D6) instead of abbreviations (RT, CV, FP). The descriptive
labels appear in parentheses in the cell drilldown and as legend
explanations only.
- **Chips**: D4 / D5 / D6 (the depth numbers operators are familiar
with)
- **Cell drilldown dimensions**: "D4 (Round Trip)", "D5 (Conversation)",
"D6 (Feature Parity)"
- **Legend**: Chip color indicators use D4/D5/D6; descriptive section
explains what each depth means with RT/CV/FP in prose
- **Tooltip**: "D4 per feature" in column tally
## Test plan
- [x] All 23 tests in cell-pieces.test.tsx and cell-drilldown.test.tsx
pass
- [x] All 15 overlay-selector-integration tests pass
- [x] 372/372 tests passing (same 3 pre-existing failures as before,
unrelated)
Partially reverts #4506: chips display depth-layer numbers (D4, D5, D6)
instead of abbreviations (RT, CV, FP). The descriptive labels Round Trip,
Conversation, and Feature Parity now appear in parentheses in the cell
drilldown dimensions and as legend explanations, keeping the abbreviations
in prose only.
- cell-pieces: chip names D4/D5/D6 (were RT/CV/FP)
- cell-drilldown: dimension labels "D4 (Round Trip)" etc.
- adaptive-legend: chip refs use D4/D5/D6, descriptive text uses RT/CV/FP
- feature-grid tooltip: D4 per feature
- tests updated to match
## Summary
Renames per-cell badge chip labels in the showcase dashboard for
clarity:
- **E2E** -> **RT** (Round Trip): single message, full-stack response
verification
- **D5** -> **CV** (Conversation): multi-turn scripted dialogue with
tool calls and content assertions
- **D6** -> **FP** (Feature Parity): cross-framework behavioral
consistency check
Reorganizes the adaptive legend to group chip color indicators first,
then adds descriptive labels explaining what RT, CV, and FP mean.
Updates cell-drilldown dimension labels, feature-grid tooltip, and all
affected tests.
## Test plan
- [x] All shell-dashboard vitest tests pass (372/372 passing; 3
pre-existing failures unrelated to this change)
- [x] No remaining references to old E2E/D5/D6 chip labels in source or
test files (D5/D6 in chips-explainer are depth layer IDs D0-D6,
correctly unchanged)
Rename per-cell badge chip labels for clarity:
- E2E -> RT (Round Trip): single message, full-stack response verification
- D5 -> CV (Conversation): multi-turn scripted dialogue with tool calls
- D6 -> FP (Feature Parity): cross-framework behavioral consistency check
Reorganize the adaptive legend to group chip color indicators first,
followed by descriptive labels explaining what RT, CV, and FP mean.
Update cell-drilldown dimension labels to match the new naming.
Update all affected tests to use the new label strings.
## Summary
- Spring-ai StreamingToolAgent now classifies tool calls as frontend vs
backend, emitting TOOL_CALL events without TOOL_CALL_RESULT for frontend
tools so the CopilotKit runtime's processAgentResult handles them
- On HITL re-invocation (tool result sent back), sends full AG-UI
message history to aimock/LLM via convertMessages() for correct fixture
matching
- Fixes 3 HITL features: hitl-text-input, hitl-approve-deny, hitl-steps
(spring-ai now 6/11 D5 green, up from 3/11)
## Remaining 5 failures (pre-existing, not caused by this PR)
| Feature | Root cause |
|---------|-----------|
| tool-rendering | CopilotKit frontend re-render drops assistant
messages (count goes 1→0) |
| shared-state-read | SharedStateReadWriteController — separate
controller, not StreamingToolAgent |
| shared-state-write | Same as shared-state-read |
| subagents | SubagentsController — aimock returns "No fixture matched"
(404) |
| mcp-apps | Depends on external MCP server (excalidraw.com) unreachable
from Docker |
All 5 fail identically on baseline (main) without this PR's changes.
## Test plan
- [x] `showcase test spring-ai --d5` — 6 green, 5 red (same 5 red on
baseline)
- [x] Direct curl to Java backend confirms correct AG-UI event sequences
for all HITL flows
- [x] No regressions: agentic-chat, gen-ui-custom, gen-ui-headless still
pass
Tyler Slaton