How real projects brief Codex, Cursor and every other agent that reads AGENTS.md.
PydanticAGENTS.mdpydantic/pydantic-ai/pydantic_ai_slim/pydantic_ai/ui/AGENTS.md
union shares (for `Message`, a string `id`), so anything else malformed must still fail. Tests for version-gated behavior belong behind `requires_ag_ui(' ')` in `tests/test_ag_ui.py`, never behind the module
PydanticAGENTS.mdpydantic/pydantic-ai/src/pydantic_ai_harness/AGENTS.md
/docs/ai/core-concepts/hooks/> - toolsets: <https://pydantic.dev/docs/ai/tools-toolsets/toolsets/> - advanced tools: <https://pydantic.dev/docs/ai/tools-toolsets/tools-advanced/> - agents: <https://pydantic.dev/docs/ai/core-concepts/agent/> - testing: <https://pydantic.dev/docs/ai/guides/testing/> 4. Inspect the installed `pydantic_ai` package source for exact hook/toolset signatures
PydanticAGENTS.mdpydantic/pydantic-ai/src/pydantic_clai2/AGENTS.md
PluginHost` instance someone constructed. No module-level dicts, no import-time side effects. Tests build a host and call `activate` directly. - **Strings are keys, never payloads.** Hook names
PydanticAGENTS.mdpydantic/pydantic-ai/tests/AGENTS.md
Testing Guidelines ## Testing philosophy The async backend defaults to asyncio. Use `uv run pytest --anyio-backend=trio --record-mode=none` to run selected portable tests on Trio without duplicating
PydanticAGENTS.mdpydantic/pydantic-ai/tests/benchmarks/AGENTS.md
performance improvement you want to implement. 2. A benchmark should be succinct, and stress tested to ensure no flakiness around
1jehuangAGENTS.md1jehuang/jcode/AGENTS.md
user asks you to review or integrate a PR or branch, you may inspect, test, and integrate that contribution regardless of author status. Do not pull in unrelated branches
teng-linAGENTS.mdteng-lin/notebooklm-py/AGENTS.md
Click), `mcp/` (FastMCP), `server/` (FastAPI), `rpc/` (batchexecute RPC facade), `_android/` (gRPC session & codecs). - **Tests & Data:** `tests/unit/`, `tests/integration/` (VCR), `tests/server/`, `tests/e2e/`, `tests/_guardrails/`, `tests/_fault_server/`, `tests/cassettes/` (`web/`, `android/`). Examples in `examples/`, tools
GoogleAGENTS.mdgoogle/gvisor/AGENTS.md
ARM64). ## Critical Development Commands AI agents should use these commands to build, test, and verify: - **Build all targets:** `make build` - **Run unit tests:** `make tests` - **Run a specific test:** `make
Hugging FaceAGENTS.mdhuggingface/trl/.ai/AGENTS.md
into **main code** and **experimental code**. * **Main code** should remain stable, consistent, and well-tested. * **Experimental code** may be less stable and may contain inconsistent patterns or limited testing. Small
nesquenaAGENTS.mdnesquena/hermes-webui/AGENTS.md
creates/uses the repo `.venv`, pins execution to Python 3.11-3.13, and installs missing dev test dependencies. `HERMES_WEBUI_TEST_PYTHON` selects the supported base interpreter used to create or rebuild
nukeopAGENTS.mdnukeop/nuclear/AGENTS.md
packages pnpm lint:fix # Lint and auto-fix pnpm type-check # TypeScript checks pnpm test # Run all tests pnpm test:coverage # Run tests with coverage pnpm clean # Clean build artifacts
VercelAGENTS.mdvercel-labs/json-render/AGENTS.md
pnpm turbo run build --filter='web^...'`, `pnpm --filter web build`, and `pnpm --filter web test:routes`. Existing-page source and Markdown parity are covered by `apps/web/tests/fixtures/docs-baseline.json`; update fixtures only when
LangChainAGENTS.mdlangchain-ai/langchainjs/AGENTS.md
Path | Description | | --------------------------- | -------------------------------- | ------------------------------------ | | `@langchain/build` | `internal/build/` | Build utilities | | `@langchain/tsconfig` | `internal/tsconfig/` | Shared TypeScript configuration | | `@langchain/standard-tests` | `libs/langchain-standard-tests/` | Standard test suite for integrations | ## Development Setup ### Prerequisites - **Node.js v24.x** (check with `node -v`) - **pnpm v10.14.0** (package
RightNow-AIAGENTS.mdRightNow-AI/openfang/AGENTS.md
Stack - typecheck: cargo build --workspace --lib - lint: cargo clippy --workspace --all-targets -- -D warnings - test: cargo test --workspace - shell: shellcheck scripts/install.sh
NVIDIAAGENTS.mdNVIDIA/Megatron-LM/AGENTS.md
Repository Guidelines ## Skills The `skills/` directory contains structured guides for common tasks (running tests, building containers, managing dependencies, submitting SLURM jobs, etc.). **Always read the relevant `SKILL.md` before starting
rowboatlabsAGENTS.mdrowboatlabs/rowboat/AGENTS.md
time and which then goes stale during co-development. Build harbor before installing or testing x: ```sh cd apps/harbor && pnpm install && pnpm -r build cd ../x && pnpm install && npm test
muratcankoylanAGENTS.mdmuratcankoylan/Agent-Skills-for-Context-Engineering/AGENTS.md
loop touches. - Live execution is the highest-signal validation for orchestration code; smoke-test changes against the actual loop before declaring them safe. - Cursor SDK is the only paid
Google CloudAGENTS.mdGoogleCloudPlatform/generative-ai/search/gemini-enterprise/ge-demo-generator/AGENTS.md
selector on English comes back in English. The selector is the control for that. - **Test language behaviour by sampling, not once.** The original bug reproduced in 2 of 5 runs
Google CloudAGENTS.mdGoogleCloudPlatform/generative-ai/search/gemini-enterprise/group-licensing/AGENTS.md
single binary: `joiner` (grant) and `garbage_collection` (revoke). ## Quick reference ```bash # Run all unit tests (no credentials required) go test ./... # Build the binary CGO_ENABLED=0 go build -trimpath -ldflags
CannerAGENTS.mdCanner/WrenAI/AGENTS.md
expected of a change here — reproduce a failure before claiming one, label honestly, make tests exercise the code they cover — and what will be closed rather than reviewed. Per-module
An open format for instructions to coding agents, read by Codex, Cursor and others. Think of it as a README written for agents.
At the repository root, with more specific files in subdirectories. Agents read the one closest to the file they're editing.
Setup and test commands, code style, and the rules a new contributor would need to know.
Claude Code reads CLAUDE.md. A one-line CLAUDE.md that points at AGENTS.md covers both.
