Files
rtk-ai__rtk/README.md
T
Florian BRUNIAUX 926e6a0dd4 feat: TOML Part 2 — user-global config, shadow warning, rtk init templates, 4 new built-in filters (#351)
* docs: update module count to 58 (toml_filter + verify_cmd)

* docs: bump version refs to 0.27.1, fix module count to 59

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

* feat: TOML DSL PR 2 — user-global config, shadow warning, init templates, 4 new filters

- toml_filter: add ~/.config/rtk/filters.toml (priority 2, between project and built-in)
- toml_filter: shadow warning when match_command overlaps a Rust-handled command
- init: rtk init generates .rtk/filters.toml template (local) and ~/.config/rtk/filters.toml (global)
- builtin_filters: add pre-commit, helm, gcloud, ansible-playbook (46/46 inline tests)
- README: add "Custom Filters" section with lookup table, primitives, examples, built-in list

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

* feat(filters): add quarto-render, pnpm-build, trunk-build from rtk discover data (52/52 tests)

* feat(filters): add docker-inspect, sops, docker-compose-ps built-in filters

Resolves issues #279, #277, #276 as TOML-native filters (no Rust required).

- docker-inspect: strip_ansi + truncate_lines_at=120 + max_lines=60
- sops: strip_ansi + strip_lines_matching blank lines + max_lines=40
- docker-compose-ps: strip_ansi + truncate_lines_at=120 + max_lines=40

6 inline tests added (2 per filter), all passing.

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

* refactor(filters): split monolithic builtin_filters.toml into 28 individual files

Replace the 848-line src/builtin_filters.toml monolith with individual
files under src/filters/<name>.toml (1 filter + tests per file).

Files are concatenated alphabetically by build.rs at compile time into
OUT_DIR/builtin_filters.toml, then embedded via a BUILTIN_TOML constant
in toml_filter.rs. The build step validates TOML syntax and detects
duplicate filter names across files.

Benefits:
- Zero merge conflicts when multiple PRs add filters (different files)
- Clear PR diffs: "+1 file of 30 lines" vs "+30 lines in 850-line file"
- Easy onboarding: copy any .toml, rename, edit 3 fields — done
- Build-time TOML validation catches syntax errors before tests run

Changes:
- build.rs (new): concat + validate src/filters/*.toml
- Cargo.toml: add [build-dependencies] toml = "0.8"
- src/filters/*.toml (28 files): split from monolith
- src/filters/README.md (new): contributor guide
- src/toml_filter.rs: BUILTIN_TOML const + 5 include_str replacements
- src/builtin_filters.toml: deleted

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

* test(filters): add 5 new unit tests for multi-file architecture + wget filter

Unit tests added (toml_filter.rs):
- test_builtin_toml_has_schema_version: ensures build.rs injects schema_version
- test_builtin_all_expected_filters_present: guards against accidental file deletion
- test_builtin_filter_count: exact count check (fails if filter added/removed without update)
- test_builtin_all_filters_have_inline_tests: prevents shipping filters with no tests
- test_new_filter_discoverable_after_concat: simulates build.rs concat, verifies lookup

New built-in filter:
- src/filters/wget.toml: compact wget download output (strips connection/resolution
  noise, short-circuits on 'saved [' to 'ok (downloaded)', 2 inline tests)

Test results: 661/661 unit tests, 60/60 inline TOML filter tests

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

* docs: add filter-workflow.md with Mermaid diagrams

Build pipeline + runtime lookup priority, both as Mermaid flowcharts.
Shows the full path from src/filters/*.toml to binary to execution.

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

* fix: remove dead TOML filters — wget, git-checkout, git-merge, git-remote, cargo-run

These filters were never reachable: Clap routes rtk git/cargo commands
to their dedicated Rust modules before run_fallback is called. The TOML
engine only fires for unknown commands. Shipping dead filters gives false
confidence (inline tests pass, but filters never activate in production).

Also fixes P1-1 from the PR #386 review: wget was in RUST_HANDLED_COMMANDS
and the shadow warning fired at compile time but did not block the build.

Removed: cargo-run.toml, git-checkout.toml, git-merge.toml,
         git-remote.toml, wget.toml

Updated test guards: count 29→24, expected list -5 names,
concat test 30→25.

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

* fix: remove 3 dead TOML filters + update guards (PR #351)

docker-inspect, docker-compose-ps, and pnpm-build are handled by
container.rs and pnpm_cmd.rs before run_fallback is reached — their
TOML filters never fire. Remove the files to avoid false documentation.

Update toml_filter.rs guards: expected list -3 names, count 24→21,
concat test 25→22. Update CONTRIBUTING.md example count to match.

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

* fix: rebase on PR #349 — apply regex anchors, make on_empty, build.rs clippy, docs macOS path

- Add (\\s|$) word-boundary anchors to 6 tofu/mix match_command regexes
  (tofu-plan, tofu-init, tofu-validate, tofu-fmt, mix-format, mix-compile)
- Add on_empty = "make: ok" to make.toml + update existing empty-output test
- Fix build.rs clippy: map_or(false, ...) → is_some_and(..)
- Add macOS alt path note to filter-workflow.md Mermaid diagram

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

---------

Co-authored-by: Claude Sonnet 4.6 <noreply@anthropic.com>
2026-03-10 11:02:18 +01:00

12 KiB

RTK - Rust Token Killer

High-performance CLI proxy that reduces LLM token consumption by 60-90%

CI Release License: MIT Discord Homebrew

Website • Install • Troubleshooting • Architecture • Discord

English • Francais • 中文 • 日本語 • 한국어 • Espanol


rtk filters and compresses command outputs before they reach your LLM context. Single Rust binary, zero dependencies, <10ms overhead.

Token Savings (30-min Claude Code Session)

Operation Frequency Standard rtk Savings
ls / tree 10x 2,000 400 -80%
cat / read 20x 40,000 12,000 -70%
grep / rg 8x 16,000 3,200 -80%
git status 10x 3,000 600 -80%
git diff 5x 10,000 2,500 -75%
git log 5x 2,500 500 -80%
git add/commit/push 8x 1,600 120 -92%
cargo test / npm test 5x 25,000 2,500 -90%
ruff check 3x 3,000 600 -80%
pytest 4x 8,000 800 -90%
go test 3x 6,000 600 -90%
docker ps 3x 900 180 -80%
Total ~118,000 ~23,900 -80%

Estimates based on medium-sized TypeScript/Rust projects. Actual savings vary by project size.

Installation

brew install rtk

Quick Install (Linux/macOS)

curl -fsSL https://raw.githubusercontent.com/rtk-ai/rtk/refs/heads/master/install.sh | sh

Installs to ~/.local/bin. Add to PATH if needed:

echo 'export PATH="$HOME/.local/bin:$PATH"' >> ~/.bashrc  # or ~/.zshrc

Cargo

cargo install --git https://github.com/rtk-ai/rtk

Pre-built Binaries

Download from releases:

  • macOS: rtk-x86_64-apple-darwin.tar.gz / rtk-aarch64-apple-darwin.tar.gz
  • Linux: rtk-x86_64-unknown-linux-musl.tar.gz / rtk-aarch64-unknown-linux-gnu.tar.gz
  • Windows: rtk-x86_64-pc-windows-msvc.zip

Verify Installation

rtk --version   # Should show "rtk 0.27.2"
rtk gain        # Should show token savings stats

Name collision warning: Another project named "rtk" (Rust Type Kit) exists on crates.io. If rtk gain fails, you have the wrong package. Use cargo install --git above instead.

Quick Start

# 1. Install hook for Claude Code (recommended)
rtk init --global
# Follow instructions to register in ~/.claude/settings.json

# 2. Restart Claude Code, then test
git status  # Automatically rewritten to rtk git status

The hook transparently rewrites commands (e.g., git status -> rtk git status) before execution. Claude never sees the rewrite, it just gets compressed output.

How It Works

  Without rtk:                                    With rtk:

  Claude  --git status-->  shell  -->  git         Claude  --git status-->  RTK  -->  git
    ^                                   |            ^                      |          |
    |        ~2,000 tokens (raw)        |            |   ~200 tokens        | filter   |
    +-----------------------------------+            +------- (filtered) ---+----------+

Four strategies applied per command type:

  1. Smart Filtering - Removes noise (comments, whitespace, boilerplate)
  2. Grouping - Aggregates similar items (files by directory, errors by type)
  3. Truncation - Keeps relevant context, cuts redundancy
  4. Deduplication - Collapses repeated log lines with counts

Commands

Files

rtk ls .                        # Token-optimized directory tree
rtk read file.rs                # Smart file reading
rtk read file.rs -l aggressive  # Signatures only (strips bodies)
rtk smart file.rs               # 2-line heuristic code summary
rtk find "*.rs" .               # Compact find results
rtk grep "pattern" .            # Grouped search results
rtk diff file1 file2            # Condensed diff

Git

rtk git status                  # Compact status
rtk git log -n 10               # One-line commits
rtk git diff                    # Condensed diff
rtk git add                     # -> "ok"
rtk git commit -m "msg"         # -> "ok abc1234"
rtk git push                    # -> "ok main"
rtk git pull                    # -> "ok 3 files +10 -2"

GitHub CLI

rtk gh pr list                  # Compact PR listing
rtk gh pr view 42               # PR details + checks
rtk gh issue list               # Compact issue listing
rtk gh run list                 # Workflow run status

Test Runners

rtk test cargo test             # Show failures only (-90%)
rtk err npm run build           # Errors/warnings only
rtk vitest run                  # Vitest compact (failures only)
rtk playwright test             # E2E results (failures only)
rtk pytest                      # Python tests (-90%)
rtk go test                     # Go tests (NDJSON, -90%)
rtk cargo test                  # Cargo tests (-90%)

Build & Lint

rtk lint                        # ESLint grouped by rule/file
rtk lint biome                  # Supports other linters
rtk tsc                         # TypeScript errors grouped by file
rtk next build                  # Next.js build compact
rtk prettier --check .          # Files needing formatting
rtk cargo build                 # Cargo build (-80%)
rtk cargo clippy                # Cargo clippy (-80%)
rtk ruff check                  # Python linting (JSON, -80%)
rtk golangci-lint run           # Go linting (JSON, -85%)

Package Managers

rtk pnpm list                   # Compact dependency tree
rtk pip list                    # Python packages (auto-detect uv)
rtk pip outdated                # Outdated packages
rtk prisma generate             # Schema generation (no ASCII art)

Containers

rtk docker ps                   # Compact container list
rtk docker images               # Compact image list
rtk docker logs <container>     # Deduplicated logs
rtk docker compose ps           # Compose services
rtk kubectl pods                # Compact pod list
rtk kubectl logs <pod>          # Deduplicated logs
rtk kubectl services            # Compact service list

Data & Analytics

rtk json config.json            # Structure without values
rtk deps                        # Dependencies summary
rtk env -f AWS                  # Filtered env vars
rtk log app.log                 # Deduplicated logs
rtk curl <url>                  # Auto-detect JSON + schema
rtk wget <url>                  # Download, strip progress bars
rtk summary <long command>      # Heuristic summary
rtk proxy <command>             # Raw passthrough + tracking

Token Savings Analytics

rtk gain                        # Summary stats
rtk gain --graph                # ASCII graph (last 30 days)
rtk gain --history              # Recent command history
rtk gain --daily                # Day-by-day breakdown
rtk gain --all --format json    # JSON export for dashboards

rtk discover                    # Find missed savings opportunities
rtk discover --all --since 7    # All projects, last 7 days

Global Flags

-u, --ultra-compact    # ASCII icons, inline format (extra token savings)
-v, --verbose          # Increase verbosity (-v, -vv, -vvv)

Examples

Directory listing:

# ls -la (45 lines, ~800 tokens)        # rtk ls (12 lines, ~150 tokens)
drwxr-xr-x  15 user staff 480 ...       my-project/
-rw-r--r--   1 user staff 1234 ...       +-- src/ (8 files)
...                                      |   +-- main.rs
                                         +-- Cargo.toml

Git operations:

# git push (15 lines, ~200 tokens)       # rtk git push (1 line, ~10 tokens)
Enumerating objects: 5, done.             ok main
Counting objects: 100% (5/5), done.
Delta compression using up to 8 threads
...

Test output:

# cargo test (200+ lines on failure)     # rtk test cargo test (~20 lines)
running 15 tests                          FAILED: 2/15 tests
test utils::test_parse ... ok               test_edge_case: assertion failed
test utils::test_format ... ok              test_overflow: panic at utils.rs:18
...

Auto-Rewrite Hook

The most effective way to use rtk. The hook transparently intercepts Bash commands and rewrites them to rtk equivalents before execution.

Result: 100% rtk adoption across all conversations and subagents, zero token overhead.

Setup

rtk init -g                 # Install hook + RTK.md (recommended)
rtk init -g --auto-patch    # Non-interactive (CI/CD)
rtk init -g --hook-only     # Hook only, no RTK.md
rtk init --show             # Verify installation

After install, restart Claude Code.

Commands Rewritten

Raw Command Rewritten To
git status/diff/log/add/commit/push/pull rtk git ...
gh pr/issue/run rtk gh ...
cargo test/build/clippy rtk cargo ...
cat/head/tail <file> rtk read <file>
rg/grep <pattern> rtk grep <pattern>
ls rtk ls
vitest/jest rtk vitest run
tsc rtk tsc
eslint/biome rtk lint
prettier rtk prettier
playwright rtk playwright
prisma rtk prisma
ruff check/format rtk ruff ...
pytest rtk pytest
pip list/install rtk pip ...
go test/build/vet rtk go ...
golangci-lint rtk golangci-lint
docker ps/images/logs rtk docker ...
kubectl get/logs rtk kubectl ...
curl rtk curl
pnpm list/outdated rtk pnpm ...

Commands already using rtk, heredocs (<<), and unrecognized commands pass through unchanged.

Configuration

Config File

~/.config/rtk/config.toml (macOS: ~/Library/Application Support/rtk/config.toml):

[tracking]
database_path = "/path/to/custom.db"  # default: ~/.local/share/rtk/history.db

[hooks]
exclude_commands = ["curl", "playwright"]  # skip rewrite for these

[tee]
enabled = true          # save raw output on failure (default: true)
mode = "failures"       # "failures", "always", or "never"
max_files = 20          # rotation limit

Tee: Full Output Recovery

When a command fails, RTK saves the full unfiltered output so the LLM can read it without re-executing:

FAILED: 2/15 tests
[full output: ~/.local/share/rtk/tee/1707753600_cargo_test.log]

Uninstall

rtk init -g --uninstall     # Remove hook, RTK.md, settings.json entry
cargo uninstall rtk          # Remove binary
brew uninstall rtk           # If installed via Homebrew

Documentation

Contributing

Contributions welcome! Please open an issue or PR on GitHub.

Join the community on Discord.

License

MIT License - see LICENSE for details.