How real projects brief Claude Code: the commands, conventions and warnings they put in CLAUDE.md.
DeepMystCLAUDE.mdDeepMyst/Mysti/CLAUDE.md
dist/extension.js` from entry point `src/extension.ts` (webpack bundles with ts-loader, target: node, CommonJS2). **Note:** Tests are not yet implemented (`npm run test` exists but has no test files). ## Development Press
ClawBioCLAUDE.mdClawBio/ClawBio/CLAUDE.md
Galaxy tools from `skills/catalog.json` (which excludes proprietary/local-only folders, so never count `SKILL.md` directories); tests from `grep -rhoE '^[[:space:]]*def test_' --include='*.py' . | wc -l`; stars, forks and contributors from
AltanSCLAUDE.mdAltanS/collie/CLAUDE.md
under `bridge/`, `cli/`, `web/src/`, `web/public/`, `scripts/`, `systemd/`, or the manifest / package files, minus the tests and hooks carved out below), you MUST, **in the same commit**, add **one bullet
manaflow-aiCLAUDE.mdmanaflow-ai/manaflow/CLAUDE.md
which supports global fetch When using try/catch, never suppress errors. Always console.error any errors. # Tests Use vitest Place test files next to the file they test using a .test.ts extension
AlexSergeyCLAUDE.mdAlexSergey/rockpack/CLAUDE.md
over `interface` - Use discriminated unions when appropriate - Prefer readonly where possible - Avoid unnecessary generics --- ## Testing Rules - Tests must be deterministic - Keep tests readable and explicit - Avoid duplicated setup logic
withkynamCLAUDE.mdwithkynam/vibecode-pro-max-kit/CLAUDE.md
agent | Explicit "ENTER EXECUTE MODE" after VALIDATE (or PLAN for trivial) | Modified source files, test results | None — explicit approval always required | | UPDATE PROCESS | vc-update-process-agent | "ENTER UPDATE PROCESS
weieskyCLAUDE.mdweiesky/cc-viewer/CLAUDE.md
must first be added to `allowBuilds` in `pnpm-workspace.yaml`. Before committing to GitHub, `pnpm run test:cli` must pass. New scripts under `packages/app/server/**` must come with unit tests in `packages/app/test/`. Update
asklokeshCLAUDE.mdasklokesh/loki-mode/CLAUDE.md
erify, every iteration. - Models: Opus for planning/architecture only; Sonnet for development and functional testing; Haiku for unit tests, monitoring, and parallelizable work. - Providers: Claude Code full (Tier 1); Cline reduced
jarrodwattsCLAUDE.mdjarrodwatts/claude-code-config/CLAUDE.md
signals (dependencies, patterns) ### State Classification: | State | Signals | Your Behavior | |-------|---------|---------------| | **Disciplined** | Consistent patterns, configs present, tests exist | Follow existing style strictly | | **Transitional** | Mixed patterns, some structure
vstorm-coCLAUDE.mdvstorm-co/pydantic-deepagents/CLAUDE.md
commit) - **Run all checks**: `make all` or `pre-commit run --all-files` - **Run tests**: `make test` - **Build docs**: `make docs` or `make docs-serve` (local development) ### Single Test Commands
Mininglamp-OSSCLAUDE.mdMininglamp-OSS/octo-server/CLAUDE.md
model) - **Default Branch**: `main` ## Common Commands ```bash # Build docker build -t octo-server . # Run tests (single module) go test ./modules/group/... go test ./modules/message/ -run TestSendMsg # Run all tests go test
jonwigginsCLAUDE.mdjonwiggins/optio/CLAUDE.md
/Applications/Android Studio.app/Contents/jbr/Contents/Home`; any JDK 17–25 works): `./gradlew assembleDebug`, `./gradlew testDebugUnitTest :core:model:test :core:network:test` (the two plain-JVM modules use `test`), `./gradlew : :recordRoborazziDebug` for screenshots (` /build/outputs/roborazzi
AgriciDanielCLAUDE.mdAgriciDaniel/banana-claude/CLAUDE.md
agents/visual-critic.md` are optional read-only specialists. They are not credential or approval boundaries. - `tests/` contains deterministic offline coverage. Public CI never calls a paid provider. ## Provider state, verified
ArcadeAICLAUDE.mdArcadeAI/arcade-mcp/CLAUDE.md
install` (runs `uv sync --extra all --extra dev` + pre-commit install) | | Run all lib tests | `make test` | | Run a single test | `uv run pytest libs/tests/core/test_toolkit.py::TestClass::test_method` | | Lint + type
DockerCLAUDE.mddocker/compose-for-agents/adk-sock-shop/CLAUDE.md
Development Tools ```bash # Type checking pyright # Code formatting and linting ruff check ruff format ``` ### Testing and Development - Access the web interface at <http://localhost:8080> after running `docker compose
beizhu-1209CLAUDE.mdbeizhu-1209/AIHelms/apps/CLAUDE.md
reload --host 0.0.0.0 --port 8000 # Format black . # Lint (with auto-fix) ruff check --fix . # Test python -m pytest -v # Targeted test python -m pytest tests/test_auth.py -v ``` ## Coding Style ### Type Annotations
beizhu-1209CLAUDE.mdbeizhu-1209/AIHelms/CLAUDE.md
celery worker (separate terminal) ./dev/start-api # Start frontend (separate terminal) ./dev/start-web ``` ## Common Commands ```bash # Backend tests (requires middleware running) cd apps && python -m pytest -v # Backend lint cd apps && black . && ruff
beizhu-1209CLAUDE.mdbeizhu-1209/AIHelms/ui/CLAUDE.md
only # Lint npm run lint # Type check npm run type-check --workspaces --if-present # Test npm test ``` ## Coding Style ### Component Structure ```vue import { ref, computed } from 'vue' import { useAuth } from '@aihelms/shared' interface Props { userId: string disabled?: boolean } const props = defineProps () const emit = defineEmits<{ delete: [userId: string] }>() const isLoading = ref(false) const displayName = computed(() => props.userId.slice(0, 8)) function handleDelete() { emit('delete', props.userId) } {{ displayName }} <button class="btn-danger" :disabled="disabled || isLoading
jezwebCLAUDE.mdjezweb/claude-skills/CLAUDE.md
directory tree in this file and the table in README.md **Categories**: `development`, `design`, `productivity`, `testing`, `security`, `database`, `monitoring`, `deployment` ## Creating a Skill See [`SKILL_SHAPE.md`](SKILL_SHAPE.md) for the canonical authoring guide
yuanchengluCLAUDE.mdyuanchenglu/claude-code-webui/CLAUDE.md
architecture, structured logging, universal Claude CLI path detection, session continuity, single binary distribution, comprehensive testing. **API Endpoints**: - `GET /api/projects` - List available project directories - `POST /api/chat` - Chat messages with streaming responses
A file Claude Code reads at the start of every session. It holds the commands, conventions and warnings the agent needs for this project.
At the repository root. Claude Code also reads CLAUDE.md files in subdirectories when it works there.
Build and test commands, the project's layout, conventions that aren't obvious from the code, and mistakes to avoid. Short files tend to work better than long ones.
Claude Code reads CLAUDE.md; most other agents read AGENTS.md. Many projects keep one and point the other at it.
