CLAUDE.md examples from real projects

AnthropicCLAUDE.md

claude-for-legal / privacy-legal

anthropics/claude-for-legal/privacy-legal/CLAUDE.md

User-specific configuration for this plugin lives at a version-independent path that survives plugin updates: ~/.claude/plugins/config/claude-for-legal/privacy-legal/CLAUDE.md Rules for every skill, command, and agent in this plugin: 1. READ configuration from that path. Not from this file. 2. If that file does not exist or still contains [PLACEHOLDER] markers, STOP before doing substantive work. Say: "This plugin needs setup before it can give you useful output. Run /privacy-legal:cold-start-interview — it takes about 10-15 minutes and every command in this plugin depends on…

No other project uses it yet

AnthropicCLAUDE.md

claude-for-legal / product-legal

anthropics/claude-for-legal/product-legal/CLAUDE.md

User-specific configuration for this plugin lives at a version-independent path that survives plugin updates: ~/.claude/plugins/config/claude-for-legal/product-legal/CLAUDE.md Rules for every skill, command, and agent in this plugin: 1. READ configuration from that path. Not from this file. 2. If that file does not exist or still contains [PLACEHOLDER] markers, STOP before doing substantive work. Say: "This plugin needs setup before it can give you useful output. Run /product-legal:cold-start-interview — it takes about 10-15 minutes and every command in this plugin depends on…

No other project uses it yet

AnthropicCLAUDE.md

claude-for-legal / regulatory-legal

anthropics/claude-for-legal/regulatory-legal/CLAUDE.md

User-specific configuration for this plugin lives at a version-independent path that survives plugin updates: ~/.claude/plugins/config/claude-for-legal/regulatory-legal/CLAUDE.md Rules for every skill, command, and agent in this plugin: 1. READ configuration from that path. Not from this file. 2. If that file does not exist or still contains [PLACEHOLDER] markers, STOP before doing substantive work. Say: "This plugin needs setup before it can give you useful output. Run /regulatory-legal:cold-start-interview — it takes about 10-15 minutes and every command in this plugin depends on…

No other project uses it yet

JetBrainsCLAUDE.md

Exposed

JetBrains/Exposed/CLAUDE.md

Exposed is a lightweight ORM framework for Kotlin that provides two APIs: - DSL API: Type-safe SQL-wrapping Domain Specific Language (in exposed-core) - Works with both JDBC (exposed-jdbc) and R2DBC (exposed-r2dbc) - DAO API: Lightweight Data Access Object API - JDBC: exposed-dao - R2DBC: exposed-dao-r2dbc (experimental preview, suspending API) Tests are organized by database and dialect. Each module has database-specific test tasks:

No other project uses it yet

GoogleCLAUDE.md

adk-go

google/adk-go/CLAUDE.md

See AGENTS.md for project context, commands, and contribution guidelines for AI coding agents.

PydanticCLAUDE.md

monty

pydantic/monty/CLAUDE.md

DON'T COMMIT UNLESS EXPLICITLY ASKED TO DO SO BY THE USER! Previous commit requests do not matter - don't commit unless you've just been explicitly asked to do so. Monty is a sandboxed Python interpreter written in Rust. It parses Python code using Ruff's ruff_python_parser but implements its own runtime execution model for safety and performance. It supports a subset of Python features. Project goals: The public data types (the MontyGraph/MontyNode value graph and the MontyObject/CallArgs/NamedValues built on it, MontyException/ExcType,…

No other project uses it yet

PydanticCLAUDE.md

monty / modules

pydantic/monty/crates/monty/src/modules/CLAUDE.md

Guidance for the Python stdlib modules implemented in this directory (re, json, math, datetime, unicodedata, …). The overriding goal is CPython parity: signatures, error messages, and error ordering must match CPython 3.14 byte-for-byte. Verify every message against CPython with the /python-playground skill before pinning it in a test. Any function beyond the trivial 0/1/2-positional shapes (covered by ArgValues::check_zero_args / get_one_arg / get_two_args / get_zero_one_arg / into_pos_only) MUST define its parameters with a #[derive(FromArgs)] struct. Never hand-roll args.into_parts() loops — they…

No other project uses it yet

AnthropicCLAUDE.md

claude-agent-sdk-python

anthropics/claude-agent-sdk-python/CLAUDE.md

Workflow jobs in this repository that call Claude run with three protections. Keep them when you add or edit a workflow. 1. Egress-firewall runner. The job has runs-on: ubuntu-24.04-firewall, a GitHub-hosted runner that filters the job's outbound network traffic. Do not move a job that calls Claude to another runner. 2. Network allow list. .github/egress-firewall.yaml lists the hosts those jobs may reach, besides any that GitHub's firewall allows by default. Keep mode: enforce, which is what makes the firewall block…

No other project uses it yet

VercelCLAUDE.md

deepsec

vercel-labs/deepsec/CLAUDE.md

Pointer file for Claude Code working in this repo. User-facing docs are in README.md; contributor docs in CONTRIBUTING.md. Read those first. Always use the technical-writer skill (.agents/skills/technical-writer/SKILL.md) when you write or edit prose for this repo, including docs, READMEs, commit messages, PR titles and descriptions, and issues. - Plugin contracts live in packages/core/src/plugin.ts. Internals route through getRegistry() from deepsec/config rather than calling organization-specific code directly. - The CLI auto-loads deepsec.config.{ts,mjs,js,cjs} from cwd upward (via packages/deepsec/src/load-config.ts, jiti). - New matchers go…

NVIDIACLAUDE.md

Isaac-GR00T

NVIDIA/Isaac-GR00T/CLAUDE.md

Isaac GR00T N1.7 is an open vision-language-action (VLA) model for generalized humanoid robot skills. The repo contains the model, training pipeline, evaluation harness, and deployment tooling. Each Jetson/Spark platform ships an activate_*.sh helper (scripts/activate_orin.sh, scripts/activate_spark.sh, scripts/activate_thor.sh) that exports platform-specific library paths. For dGPU, the standard source .venv/bin/activate is sufficient.

AnthropicCLAUDE.md

defending-code-reference-harness

anthropics/defending-code-reference-harness/CLAUDE.md

This repo has two halves: - Interactive skills (.claude/skills/) — read and write files in the repo (no target-code execution, except /dnr-hunt / /dnr-respond, which run the bundled demo app…

Model Context ProtocolCLAUDE.md

registry

modelcontextprotocol/registry/CLAUDE.md

Import @README.md Publishing an MCP server does NOT happen through a pull request to this repository. Do NOT create a PR that adds a server.json or a server directory anywhere…

SentryCLAUDE.md

MobileBuildMCP

getsentry/MobileBuildMCP/CLAUDE.md

No summary in the file. Open it to read it.

Microsoft AzureCLAUDE.md

Azure-Sentinel

Azure/Azure-Sentinel/CLAUDE.md

For Microsoft Sentinel-to-Defender-XDR migration or authoring work, first read: Tools\SentinelToXDRMigration\agent-workflows\end-to-end-migration.md Follow that canonical workflow and use the repository CLI and PowerShell entry points directly. Do not replace deterministic conversion, validation, packaging, deployment, or parity logic with model-generated approximations. Default to qualification, with --workflow-profile authoring available as an explicit offline/ISV fallback when no live lab is available. Use V3.1 through package-v3-1 for Sentinel and XDR solution packages. Resume from workflow-state.json, preserve disabled detections, and request explicit approval before writes.

ElasticCLAUDE.md

go-elasticsearch

elastic/go-elasticsearch/CLAUDE.md

No summary in the file. Open it to read it.

VercelCLAUDE.md

streamdown / website

vercel/streamdown/apps/website/.claude/CLAUDE.md

This project uses Ultracite, a zero-config Biome preset that enforces strict code quality standards through automated formatting and linting. Biome (the underlying engine) provides extremely fast Rust-based linting and formatting.…

No other project uses it yet

GitHubCLAUDE.md

gh-aw / engines

github/gh-aw/pkg/workflow/data/engines/claude.md

Shared engine configuration for Claude Code. -->

NVIDIACLAUDE.md

TensorRT-Model-Optimizer / debugger

NVIDIA/TensorRT-Model-Optimizer/tools/debugger/CLAUDE.md

This directory contains a file-based command relay for executing commands inside a remote Docker container from the host machine (where Claude Code runs). The user must start the server inside Docker first: Then Claude Code performs the handshake:

NVIDIACLAUDE.md

TensorRT-Model-Optimizer / launcher

NVIDIA/TensorRT-Model-Optimizer/tools/launcher/CLAUDE.md

The launcher submits ModelOpt quantization, training, and evaluation jobs to Slurm clusters or runs them locally with Docker. The --yaml format maps top-level keys to launch() function arguments: Key conventions: 65 unit tests in tests/. Run standalone without installing modelopt: From the launcher directory: Tests cover: core dataclasses, factory registry, global_vars interpolation, YAML formats, Docker/Slurm executor construction (mocked), environment merging, metadata writing, and end-to-end Docker launch via subprocess.

CloudflareCLAUDE.md

vibesdk

cloudflare/vibesdk/CLAUDE.md

VibeSDK is an agentic full-stack application builder on Cloudflare. Tech Stack: - Frontend: React 19, TypeScript, Vite, TailwindCSS, React Router v7 - Backend: Cloudflare Workers, Durable Objects, Hono, D1, R2, and KV - Agent: Cloudflare Think with AI Gateway model routing - Workspace: SpaceDO Durable Objects - Version history: Cloudflare Artifacts - Preview runtime: Worker Loader bindings and Dynamic Workers - Generated app data: Durable Object Facets with isolated SQLite - WebSocket: PartySocket for real-time communication Project Structure: - /src…