How real projects brief Claude Code: the commands, conventions and warnings they put in CLAUDE.md.
vibe-logCLAUDE.mdvibe-log/vibe-log-cli/CLAUDE.md
dist/ # Compiled JavaScript output ├── src/ │ ├── commands/ # CLI command implementations │ ├── lib/ # Core business logic │ │ ├── api-client.ts # Secure API communication │ │ ├── auth/ # Authentication modules │ │ ├── readers/ # Session data readers │ │ ├── sub-agents/ # Claude Code sub-agent
TheDecipheristCLAUDE.mdTheDecipherist/claude-code-mastery-project-starter-kit/CLAUDE.md
Never copied to scaffolded projects. **Project commands:** `help`, `review`, `commit`, `progress`, `test-plan`, `architecture`, `security-check`, `optimize-docker`, `create-e2e`, `create-api`, `worktree`, `refactor`, `diagram`, `setup`, `what
TheDecipheristCLAUDE.mdTheDecipherist/claude-code-mastery-project-starter-kit/global-claude-md/CLAUDE.md
Global CLAUDE.md — Security Gatekeeper & Standards > Place this at ~/.claude/CLAUDE.md > It applies to EVERY project you work on. > Based on Claude Code Mastery Guides V1-V5 by TheDecipherist --- ## Identity - GitHub: **YourUsername
agents-universeCLAUDE.mdagents-universe/agents-universe/CLAUDE.md
docker-compose.example.yml docker-compose.yml) docker compose up docker compose up --build ``` ## Key Conventions 1. **Token security** — Encrypted tokens are never logged, printed, or included in error messages. AES-256-GCM in `token_vault.py
DatabricksCLAUDE.mddatabricks/databricks-agent-skills/CLAUDE.md
such as Anthropic's official directory, is additive and never replaces the entry here. ## Security When documenting examples, obfuscate sensitive info: - Workspace IDs: use `1111111111111111` not real IDs - URLs
claude-did-thisCLAUDE.mdclaude-did-this/MCPControl/CLAUDE.md
tests with Claude and MCPControl ## Running with HTTPS/TLS MCPControl supports HTTPS for secure SSE connections (mandatory per MCP spec for production): - `node build/index.js --sse --https --cert /path/to/cert.pem --key /path/to/key.pem` - Default
niaka3dayoCLAUDE.mdniaka3dayo/agent-skills-vrc-udon/CLAUDE.md
true` on `main` - PR required for all changes ### Branch Naming `feature/*`, `fix/*`, `docs/*`, `refactor/*`, `security/*`, `chore/*` ## Key Files | File | Purpose | |------|---------| | `package.json` | npm metadata, `files` array controls what gets published | | `skills
niaka3dayoCLAUDE.mdniaka3dayo/agent-skills-vrc-udon/templates/CLAUDE.md
authorize the caller separately from receiver ownership, and never use instance master as a security or access-control boundary. ## Skills | Skill | Purpose | Path | |-------|---------|------| | `unity-vrc-udon-sharp` | UdonSharp coding, networking
sepivipCLAUDE.mdsepivip/SeekerClaw/CLAUDE.md
app/src/main/assets/nodejs-project/mcp-client.js` (MCPClient + MCPManager) - Integration: `main.js` merges MCP tools into TOOLS array, routes `mcp__ __ ` calls - Security: descriptions sanitized, SHA-256 rug-pull detection, results wrapped as untrusted content - Rate limiting: 10/min
knoxgraemeCLAUDE.mdknoxgraeme/skillfish/CLAUDE.md
JSON modes (--json flag). JSON output includes structured exit_code and error arrays. **Security**: Path validation prevents directory traversal attacks. Name validation (`isValidName()` in constants.ts) ensures owner/repo names only contain
PackmindHubCLAUDE.mdPackmindHub/packmind/CLAUDE.md
strategy - Each sub-task should have its own commit (as per Task splitting section) ## Security - **Secrets Detection**: GitGuardian runs in CI to detect leaked secrets - **Secrets Retrieval**: Always use `Configuration.getConfig
kiliczshCLAUDE.mdkiliczsh/claude-cmd/CLAUDE.md
manager class: - `CommandManager` - Command installation, search, deletion - `ClaudeMdManager` - CLAUDE.md file management - `PermissionsManager` - Security and permissions - `MCPManager` - Model Context Protocol servers - `WorkflowManager` - Command workflows - `SettingsManager` - Configuration management - `ProjectManager` - Project initialization - `HelpManager
ThreeFish-AICLAUDE.mdThreeFish-AI/analysis_claude_code/CLAUDE.md
Intelligent Context Management** - 92% threshold auto-compression with wU2 compressor 4. **6-Layer Security Framework** - From UI validation to execution sandboxing ### Key Components Analyzed - **Agent Loop System** - Asynchronous generator-based
SaynaAICLAUDE.mdSaynaAI/sayna/CLAUDE.md
codebase includes detailed development rules in `.cursor/rules/`: - **`rust.mdc`**: Rust best practices, design patterns, performance, security, testing - **`core.mdc`**: STT/TTS provider abstractions (`BaseSTT`, `TTSProvider` traits) - **`axum.mdc`**: Axum framework patterns for WebSocket
gendigitalincCLAUDE.mdgendigitalinc/sage/CLAUDE.md
Plugin Config** — `hooks/hooks.json` registers hooks, `.claude-plugin/plugin.json` is the plugin manifest, `skills/security-awareness/SKILL.md` provides security knowledge to Claude. **Data flow (Claude Code):** `mcp_tool` hook call → `sage_claude_pre_tool
ucsandmanCLAUDE.mducsandman/DashClaw/CLAUDE.md
file edits) belong on a cheap model (Haiku/Sonnet); reserve Opus/Fable for architecture, debugging, and security reasoning. ## Gotchas (what you can't infer from the code) - **After pulling changes that touch
FradSerCLAUDE.mdFradSer/mcp-server-mas-sequential-thinking/CLAUDE.md
unit # Run unit tests only make test-integration # Run integration tests only make test-security # Run security tests only make test-config # Run configuration tests make test-fast # Fast
Prorise-coolCLAUDE.mdProrise-cool/Claude-Code-Multi-Agent/CLAUDE.md
激活 `server-sequential-thinking` 分析根因 ➔ 修复。 ### 5. [MODE: REVIEW] (审查与验收) **视角**: ⚖️ 质量保证 (QA) + 安全审计 (Security) * **目标**: 确保零偏差，代码符合高标准。 * **禁止**: 掩盖问题、忽略边缘情况。 * **强制动作序列**: 1. **偏差比对**: 使用 `github` diff 或 `filesystem` 读取代码
theexperiencecompanyCLAUDE.mdtheexperiencecompany/gaia/.github/CLAUDE.md
Quality") — hygiene.** Twenty lanes (Biome, tsc, `python-static` = ruff + custom AST lints + complexity + docstrings + security in one job, mypy, dead code, evlog-map observability score, wide-event cross-runtime conformance
AliAkhtari78CLAUDE.mdAliAkhtari78/SpotifyScraper/CLAUDE.md
never need `SPOTIFY_SP_DC`); never edit or weaken a gate, `.github/`, release, or security config; make the smallest correct change; and the **evidence standard** is the bar for "done
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.
