How real projects brief Claude Code: the commands, conventions and warnings they put in CLAUDE.md.
claude-code-bestCLAUDE.mdclaude-code-best/claude-code/CLAUDE.md
Protocol (Agent Client Protocol) - **`src/services/acp/`** — ACP agent 实现，包含 `agent.ts`（AcpAgent 类）、`bridge.ts`（Claude Code ↔ ACP 桥接）、`permissions.ts`（权限处理）、`entry.ts`（入口）。 - **`packages/acp-link/`** — ACP 代理服务器，将 WebSocket 客户端桥接到 ACP agent。提供
steipeteCLAUDE.mdsteipete/CodexBar/docs/claude.md
usage/status parsing - Updating Claude OAuth/web endpoints or cookie import - Adjusting Claude CLI PTY automation - Reviewing local cost usage scanning --- # Claude provider Claude supports three usage data paths plus local cost
teng-linCLAUDE.mdteng-lin/notebooklm-py/CLAUDE.md
Trigger Claude review by commenting `@claude review` on the PR. 3. Address every review comment (`gemini-code-assist`, `coderabbitai`, `claude[bot]`): fix, push, reply on the thread (`Addressed
emcie-coCLAUDE.mdemcie-co/parlant/CLAUDE.md
again until you get it. 4. Implement the tests first. Ask for confirmation and code review. 5. Once tests are approved, once again suggest your implementation plan for making them
muratcankoylanCLAUDE.mdmuratcankoylan/Agent-Skills-for-Context-Engineering/CLAUDE.md
skills) - `SKILL.md` (root) - Collection-level metadata and skill map - `.claude-plugin/marketplace.json` - Claude Code marketplace manifest (single bundled plugin, v2.5.0) - `.plugin/plugin.json` - Open Plugins format manifest (v2.5.0) ## Build & Test Commands
AgriciDanielCLAUDE.mdAgriciDaniel/claude-seo/CLAUDE.md
Analysis Skill ## Project Overview This repository contains **Claude SEO**, a Tier 4 Claude Code skill for comprehensive SEO analysis across all industries. It follows the Agent Skills open standard
CannerCLAUDE.mdCanner/WrenAI/.claude/CLAUDE.md
downstream. - Every assertion must be capable of failing. An assertion about a condition the code cannot produce is noise. - Confirm the test actually runs in CI. Some test files
AnthropicCLAUDE.mdanthropics/claude-quickstarts/managed-agents/assistant-ui/CLAUDE.md
multiagent: { type: "coordinator", agents: [...] }` roster so the analyst hands its numbers to a toolless reviewer. The handoff events (`session.thread_created`, `agent.thread_message_sent/received`) already arrive on the stream; the reducer
img2threejsCLAUDE.mdimg2threejs/img2threejs/CLAUDE.md
first. Do not infer visual evidence from runtime readiness, structural tests, inline previews, or code review, and do not claim the visual gate passed. **Scope of this gate.** It governs
open-metadataCLAUDE.mdopen-metadata/OpenMetadata/CLAUDE.md
endpoints, Playwright for UI | | `verification` | before claiming "done" — run real commands, show evidence | | `code-review` | reviewing a diff/PR — spec compliance then code quality | | `java-checkstyle` | after touching `.java` — runs
yusufkaraaslanCLAUDE.mdyusufkaraaslan/Skill_Seekers/CLAUDE.md
ignore=tests/test_mcp_server.py --ignore=tests/test_install_skill_e2e.py -q # Single test pytest tests/test_scraper_features.py::test_detect_language -vv -s # Code quality (must pass before push - matches CI) uvx ruff check src/ tests/ uvx ruff format
coleam00CLAUDE.mdcoleam00/context-engineering-intro/claude-code-full-guide/CLAUDE.md
names** before use - **Keep CLAUDE.md updated** when adding new patterns or dependencies - **Test your code** - No feature is complete without tests - **Document your decisions** - Future developers (including yourself) will thank
Orchestra-ResearchCLAUDE.mdOrchestra-Research/AI-Research-SKILLs/CLAUDE.md
idea to paper. Each skill provides expert-level guidance (200-500 lines) with real code examples, troubleshooting guides, and production-ready workflows. **Mission**: Enable AI agents to autonomously conduct
MemTensorCLAUDE.mdMemTensor/MemOS/CLAUDE.md
evidence | | `design-reviewer` | Read-only | Review design docs (architecture / interface / performance / security / requirement coverage) | | `code-reviewer` | Read-only | Review diffs and return APPROVE or CHANGES_REQUESTED | | `backend-dev` | Read-write
JeffallanCLAUDE.mdJeffallan/claude-skills/CLAUDE.md
searchable keywords - `role`: `specialist` | `expert` | `architect` | `engineer` - `scope`: `implementation` | `review` | `design` | `system-design` | `testing` | `analysis` | `infrastructure` | `optimization` | `architecture` - `output-format`: `code` | `document` | `report` | `architecture` | `specification` | `schema` | `manifests` | `analysis` | `analysis
ykdojoCLAUDE.mdykdojo/claude-code-tips/CLAUDE.md
Draft it to a temp file and open it in VS Code (`code `) for me to review before filing - never file without showing me first. - File from host
frankbriaCLAUDE.mdfrankbria/ralph-claude-code/CLAUDE.md
write-file`/`kill`; JSON on stdout, `exec` streams remote output + propagates the remote exit code; secrets only via env/stdin, never argv — tests substitute the interpreter via `SANDBOX_E2B_PYTHON
AnthropicCLAUDE.mdanthropics/claude-for-legal/ai-governance-legal/CLAUDE.md
distribution, category breakdown, or timeline as fits), and the reviewer note carried over. In Cowork this renders inline. In Claude Code I'll write an HTML file to [outputs folder
AnthropicCLAUDE.mdanthropics/claude-for-legal/commercial-legal/CLAUDE.md
distribution, category breakdown, or timeline as fits), and the reviewer note carried over. In Cowork this renders inline. In Claude Code I'll write an HTML file to [outputs folder
AnthropicCLAUDE.mdanthropics/claude-for-legal/corporate-legal/CLAUDE.md
distribution, category breakdown, or timeline as fits), and the reviewer note carried over. In Cowork this renders inline. In Claude Code I'll write an HTML file to [outputs folder
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.
