How real projects brief Claude Code: the commands, conventions and warnings they put in CLAUDE.md.
GitHubCLAUDE.mdgithub/gh-aw-firewall/CLAUDE.md
containers/enclave/` - Enabled by declaring `enclaves` entries in the AWF config file - One AWF-owned MCP server (`enclave-mcp-server`) exposes enabled enclave executors only through compiler-launched `gh-aw-mcpg
touwaeriolCLAUDE.mdtouwaeriol/claude-code-plus/CLAUDE.md
路径 | 职责 | |------|------|------| | **frontend** | `frontend/` | Vue 3 前端，聊天界面、工具展示 | | **ai-agent-server** | `ai-agent-server/` | HTTP/RSocket 服务器，连接前端与 SDK | | **claude-agent-sdk** | `claude-agent-sdk/` | Claude CLI 封装，包含
AWSCLAUDE.mdawslabs/visual-asset-management-system/CLAUDE.md
value and its assertion) 6. `tools/VamsMCP/pyproject.toml` (`[project]` `version`) 7. `tools/VamsMCP/vams_mcp/__init__.py` (`__version__` — what the running MCP server reports; Pattern 7 rule 6) The sub-package `package.json`/`package-lock.json` files under `infra
iamtouchskyerCLAUDE.mdiamtouchskyer/memex/CLAUDE.md
agents. Markdown cards in `~/.memex/cards/` with `[[wikilinks]]`. npm package `@touchskyer/memex`. Distributed as CLI + MCP server + Claude Code plugin + VS Code extension + Pi extension. ## Before You Touch Anything ```bash npm test
NikiforovAllCLAUDE.mdNikiforovAll/claude-code-rules/CLAUDE.md
skill) │ ├── agents/ # Agent definitions (.md) │ ├── commands/ # Slash commands (.md) │ ├── hooks/hooks.json # Event hooks │ ├── .mcp.json # MCP servers │ └── .lsp.json # LSP servers ├── .claude-plugin/ │ └── marketplace.json # Catalogs all plugins (marketplace: "cc-handbook") ├── website/ # Docusaurus documentation
UiPathCLAUDE.mdUiPath/coder_eval/CLAUDE.md
Task Definition Guide](docs/TASK_DEFINITION_GUIDE.md) § `stop_early`. - **Sandbox isolation**: tasks that don't need MCP servers should set `setting_sources: []` in their `agent:` block, to isolate the sandbox from the host
foxj77CLAUDE.mdfoxj77/claudectx/CLAUDE.md
manages configuration profiles stored in `~/.claude/profiles/`, enabling switching between different settings, CLAUDE.md files, and MCP servers. ## Build & Test Commands ```bash make build # Build binary (outputs ./claudectx) make install-user # Install
shivasuryaCLAUDE.mdshivasurya/code-pathfinder/CLAUDE.md
output ``` ### Linting ```bash gradle lintGo # or directly: golangci-lint run ``` ### Running the Binary ```bash # MCP server mode (for AI coding assistants) ./build/go/pathfinder serve --project ./build/go/pathfinder serve --http --address :8080 --project
TikiTribeCLAUDE.mdTikiTribe/claude-secure-coding-rules/CLAUDE.md
projects) │ │ ├── owasp-2025.md # OWASP Top 10 2025 security rules │ │ ├── mcp-security.md # Model Context Protocol (MCP) security rules │ │ ├── ai-security.md # AI/ML system security rules │ │ └── agent-security.md # Agentic AI security rules │ │ │ ├── languages/ # Language-specific security rules
TaoidleCLAUDE.mdTaoidle/plan-cascade/CLAUDE.md
executable tasks with multi-agent collaboration. It operates as a Claude Code Plugin (stable), MCP Server (stable), Standalone CLI (in development), and Desktop App (in development). ## Build & Development Commands **Requires
TaoidleCLAUDE.mdTaoidle/plan-cascade/desktop/CLAUDE.md
task execution with features like agent management, analytics, quality gates, git worktrees, timeline checkpoints, MCP server integration, knowledge base (RAG), plugins, guardrails, webhooks, remote control, and A2A protocol. ## Build & Development
Dave-LondonCLAUDE.mdDave-London/Pare/rules/CLAUDE.md
Pare MCP Tools — Claude Code Instructions Pare is a suite of MCP servers that wrap CLI tools and return structured JSON, saving tokens and improving reliability. ## Core Rule Always prefer
snowmeadCLAUDE.mdsnowmead/rust-docs-mcp/CLAUDE.md
When testing the rust-docs MCP server, call the tools from the rust-docs-testing MCP. ## MCP Rust SDK Macros ### Core Macros for Tool Declaration 1. **`#[tool_router]`** - Applied
Weebapp003CLAUDE.mdWeebapp003/tradingview-mcp/CLAUDE.md
study by default (pass `max_labels` to override) ## Architecture ``` Claude Code ←→ MCP Server (stdio) ←→ CDP (localhost:9222) ←→ TradingView Desktop (Electron) ``` Pine graphics path: `study._graphics._primitivesCollection.dwglines.get('lines').get(false)._primitivesDataById
wanaku-aiCLAUDE.mdwanaku-ai/wanaku/CLAUDE.md
persistence, registry impl), `filters/` (core MCP filters), `features/` (self-contained feature crates: mcp-metadata, evaluator, intercept, chat), `server/` (binary, pipeline, mgmt API), `ui/admin/` (React embedded UI). Feature crates that only
deepnoodle-aiCLAUDE.mddeepnoodle-ai/dive/CLAUDE.md
client adapter using the official `a2a-go/v2` SDK (separate Go module: `github.com/deepnoodle-ai/dive/a2a`). `Server` exposes a Dive agent as an A2A endpoint (JSON-RPC or REST). `RemoteAgent` calls remote
chaohong-aiCLAUDE.mdchaohong-ai/ai-auto-work/CLAUDE.md
Engine/` → 读取 `.ai/constitution/engine.md`；不读取 GameServer 文档 - 只改 `GameServer/` → 读取 `GameServer/CLAUDE.md`（通用入口），再根据命中的子服务读取对应 `servers/ /CLAUDE.md`；**不读取** Backend / MCP / Engine / Frontend 的宪法和生命周期文档（GameServer 是独立领域，技术栈和架构模式与创作平台完全不同） - 只有同时出现跨目录耦合、跨模块接口变更、端到端链路变更时，才读取系统级架构文档 - **宪法详细规则**：命中特定场景时读取对应子文件 - Backend
jdilla1277CLAUDE.mdjdilla1277/agentcad/CLAUDE.md
agentcad agentcad is a local CLI and MCP server for AI-agent CAD workflows. ## Repo Identity This checkout is the **public** `jdilla1277/agentcad` repo. It is for externally safe package code
kirschbaum-developmentCLAUDE.mdkirschbaum-development/laravel-loop/CLAUDE.md
code in this repository. ## Laravel Loop Package Laravel Loop is a Model Context Protocol (MCP) server package for Laravel applications that enables AI assistants to interact with Laravel apps through
psu-efdCLAUDE.mdpsu-efd/pyHMT2D/CLAUDE.md
both models. ### AI Tools (`pyHMT2D/AI_Tools/`) 27 AI-agent-callable tools exposed as an MCP server. Self-contained directory. ``` AI_Tools/ ├── mcp_server.py # FastMCP server entry point (hmt-mcp-server) ├── state.py # HydraulicSession
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.
