How real projects brief Codex, Cursor and every other agent that reads AGENTS.md.
steipeteAGENTS.mdsteipete/CodexBar/AGENTS.md
handoff; add focused `swift test --filter ...` runs for parser/provider fixes when possible. - After any code change, run `make check` and fix all reported format/lint issues before handoff. - Prefer CLI/focused tests
MicrosoftAGENTS.mdmicrosoft/onnxruntime/AGENTS.md
github/skills/). Load each skill whose description matches the requested task, subsystem, or behavior. For code reviews, follow the `/code-review` skill in addition to relevant domain skills and matching path-scoped
Hugging FaceAGENTS.mdhuggingface/peft/.ai/AGENTS.md
Pure code-agent PRs are not allowed: a human submitter must understand and be able to defend the change end-to-end. - The submitting human is responsible for reviewing every
GitHubAGENTS.mdgithub/docs/content/copilot/responsible-use/agents.md
GitHub Copilot includes several agentic features that go beyond suggestion and conversation—they can review code, take action on your behalf, and build applications. This card covers the following experiences
JetBrainsAGENTS.mdJetBrains/intellij-community/AGENTS.md
order. - User-visible strings belong in `*.properties` for localization. ## Language - Read [Code Style](./.agents/skills/code-style/SKILL.md) before writing or reviewing Kotlin or Java. ## Writing Write every user-visible artifact in [ASD-STE100
GoogleAGENTS.mdgoogle/filament/AGENTS.md
Synchronization guidelines for Java/JNI (Android) and JavaScript/TypeScript (Web/WASM) when public C++ APIs change. ### 6. Code Review & Invariant Auditing * [skills/code_review/SKILL.md](skills/code_review/SKILL.md): Adversarial systems, concurrency, memory-model, and architecture code review protocol
MicrosoftAGENTS.mdmicrosoft/fluentui/AGENTS.md
copy these) - **DO NOT copy patterns from `packages/react/` (v8).** That's maintenance-only legacy code using runtime styling, class components, and different APIs. - **DO NOT use `@fluentui/react` imports
kortix-aiAGENTS.mdkortix-ai/suna/AGENTS.md
ponytail** skill at level `full`. Load it before you write, fix, refactor, or review code, and before you add a dependency. Level switch: `/ponytail lite|full|ultra`. Off: "stop ponytail
PydanticAGENTS.mdpydantic/pydantic-ai/AGENTS.md
commit/CI gates cover them at the right time. # Coding Guidelines When generating or reviewing code anywhere in this repo, always read [agent_docs/index.md](agent_docs/index.md) and follow/enforce those guidelines
PydanticAGENTS.mdpydantic/pydantic-ai/docs/AGENTS.md
fence, such as `{test="skip" lint="skip"}`, rather than adding tooling suppressions to pedagogical code. Examples that are linted are also type-checked with pyright; opt one out with `typecheck
PydanticAGENTS.mdpydantic/pydantic-ai/.github/workflows/AGENTS.md
edit from outside the org is a supply-chain change rather than an ordinary code review — a maintainer carries legitimate ones forward in their own PR (#7024 is the case
PydanticAGENTS.mdpydantic/pydantic-ai/pydantic_ai_slim/pydantic_ai/durable_exec/AGENTS.md
implementation rename unless the migration of in-flight executions is intentional and reviewed. Resolve base and per-tool configuration with `OperationConfigRole` and `DurableOperationId`. Handle known ID variants and retain
PydanticAGENTS.mdpydantic/pydantic-ai/src/pydantic_ai_harness/AGENTS.md
state-machine-based workflow that drives AICA through phases (TRIAGE -> GOALS -> PLAN -> CODE -> VERIFY -> REVIEW -> PUBLISH) - **DDD+ protocol**: classification system for PR review comments (do, dismiss, discuss, waiting, done) ## AICA
PydanticAGENTS.mdpydantic/pydantic-ai/src/pydantic_clai2/AGENTS.md
must equal `Hooks.on`'s attribute names plus the host names. Drift fails CI, not code review. 3. Fire it from exactly one place in the shell. 4. Document
PydanticAGENTS.mdpydantic/pydantic-ai/tests/AGENTS.md
make assertions about the HTTP requests if needed. To assert what the code actually sent — headers included — take the `request_capture` fixture instead. It reads the wire rather than
1jehuangAGENTS.md1jehuang/jcode/AGENTS.md
your own branch and preserve unrelated work. When the user asks you to review or integrate a PR or branch, you may inspect, test, and integrate that contribution regardless
Hugging FaceAGENTS.mdhuggingface/trl/.ai/AGENTS.md
must also add a corresponding subsection to `paper_index.md`. When reviewing such PRs, ensure that `paper_index.md` was updated. ### Code duplication and consistency Trainers in this repository are **self-contained by design
nesquenaAGENTS.mdnesquena/hermes-webui/AGENTS.md
behavior, setup, usage, and the docs index - `CONTRIBUTING.md` for contribution scope, PR format, and review requirements - `docs/CONTRACTS.md` to find the contract/RFC for the subsystem being changed - `docs/GUIDELINES.md` for bug-class
NVIDIAAGENTS.mdNVIDIA/Megatron-LM/AGENTS.md
apply it to `megatron/training` or other training-loop code unless the PR explicitly opts into that migration. - In reviews, flag new direct `parallel_state.get_*_group()` usage in `megatron/core` unless
muratcankoylanAGENTS.mdmuratcankoylan/Agent-Skills-for-Context-Engineering/AGENTS.md
encyclopedia backbone. Promotion is gated by `research_loop.py promote-mechanisms` with a recorded reviewer; ledgers live under `researcher/mechanisms/ledgers/`. - Claim provenance for numeric or volatile claims lives in `researcher/claims/index.jsonl`. Add an entry
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.
