AGENTS.md examples from real projects

gitbutlerappAGENTS.md

gitbutler

gitbutlerapp/gitbutler/AGENTS.md

reviewable changes; avoid unrelated rewrites. - Use the simplest design that solves the actual problem; do not add speculative machinery, and remove machinery your change makes unnecessary. - Inspect nearby code before

No other project uses it yet

gitbutlerappAGENTS.md

gitbutler / crates

gitbutlerapp/gitbutler/crates/AGENTS.md

snapshot-based; do not design bespoke inverse operations for it. - Before changing or reviewing code that derives graph/workspace/branch/stack/commit relationships, reachability, dependencies, ordering, operation targets, or Git graph/history/ref-placement mutations, use `crates/WORKSPACE_MODEL.md

No other project uses it yet

Hugging FaceAGENTS.md

peft

huggingface/peft/.ai/AGENTS.md

read `tests/README.md` for detailed instructions on testing in PEFT. Test code that is improperly implemented imposes a big review burden on maintainers. #### Test selection If you make a testable change

netbox-communityReads credentialsAGENTS.md

netbox

netbox-community/netbox/AGENTS.md

short: - AI tools may be used to explore the codebase and to generate code. - The contributor must fully understand, and be able to explain, every change submitted. - Issues, pull request

oracleAGENTS.md

graal

oracle/graal/docs/AGENTS.md

Graal repo. Ask your agent to consult it before drafting, rewriting, or reviewing documentation so it can follow GraalVM documentation conventions for terminology, tone, spelling, and Markdown structure. Apply these

oracleAGENTS.md

graal / substratevm

oracle/graal/substratevm/AGENTS.md

suffices; regex support increases image code size. - When porting HotSpot code, preserve its names and structure where practical so reviewers and maintainers can compare the implementations. - If you touch documented

1jehuangAGENTS.md

jcode

1jehuang/jcode/AGENTS.md

finish and integrate safely rather than switching blindly. When the user asks you to review or integrate a PR or branch, you may inspect, test, and integrate that contribution regardless

No other project uses it yet

containerdAGENTS.md

containerd

containerd/containerd/AGENTS.md

guide's [Coding Agent Usage](https://github.com/containerd/project/blob/main/CONTRIBUTING.md#coding-agent-usage) policy and the repository's [submission requirements](CONTRIBUTING.md#automated-and-ai-generated-contributions). The human contributor reviews all content before submission

xbmcAGENTS.md

xbmc

xbmc/xbmc/AGENTS.md

free simply because it is easy to produce. Optimize for code that needs to exist and can be reviewed and maintained. ## Release stabilization When working on a release branch

GitHubAGENTS.md

docs / responsible-use

github/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

walkinglabsAGENTS.md

learn-harness-engineering / templates

walkinglabs/learn-harness-engineering/skills/harness-creator/templates/agents.md

AGENT_FILE_NAME}} {{PROJECT_PURPOSE}} ## Startup Workflow Before writing code: 1. **Confirm working directory** with `pwd` 2. **Read this file** completely 3. **Read project docs if present** (`docs/ARCHITECTURE.md`, `docs/PRODUCT.md`, README

NVIDIAReads credentialsAGENTS.md

OpenShell

NVIDIA/OpenShell/AGENTS.md

first. We design systems and use agents to implement them — this is not vibe coding. The product provides safe, sandboxed runtimes for autonomous AI agents, and the project itself

No other project uses it yet

wekanAGENTS.md

wekan

wekan/wekan/AGENTS.md

section links a commit that release does not contain. The contributor's half is review. Nobody but the maintainer commits to `main` in wekan/wekan, so a change from anyone else

PydanticAGENTS.md

pydantic-ai

pydantic/pydantic-ai/AGENTS.md

slow and CI covers 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.md

pydantic-ai

pydantic/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

PydanticReads credentialsAGENTS.md

pydantic-ai / workflows

pydantic/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.md

pydantic-ai / durable_exec

pydantic/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.md

pydantic-ai / pydantic_ai_harness

pydantic/pydantic-ai/src/pydantic_ai_harness/AGENTS.md

collection of pre-made capabilities for Pydantic AI. - **AICA**: AI Code Assistant -- the automated agent that implements issues, reviews plans, and handles PR feedback ## AICA preflight Before implementing or reviewing

PydanticReads credentialsAGENTS.md

pydantic-ai / tests

pydantic/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