diffmode_free
acogood/diffmode_free/CLAUDE.md
backend. Perplexity is opt-in.** Nothing detects, probes for, or announces a Perplexity MCP; a configured server is not consent. It is dispatched only when the founder names
How real projects brief Claude Code: the commands, conventions and warnings they put in CLAUDE.md.
acogood/diffmode_free/CLAUDE.md
backend. Perplexity is opt-in.** Nothing detects, probes for, or announces a Perplexity MCP; a configured server is not consent. It is dispatched only when the founder names
oxgeneral/ORCH/CLAUDE.md
names (e.g. `qa`, `review`, `ship`) — Markdown content injected into agent prompts at dispatch time - **MCP skills**: colon-separated names (e.g. `testing-suite:generate-tests`) — handled natively by Claude CLI, skipped
Theopsguide/gemini-code-flow/CLAUDE.md
patterns for Gemini's API - Leverage Gemini's multimodal capabilities - Integrate with Gemini's MCP server support - Utilize Gemini's large context window (1M tokens) ## Development Approach Following SPARC methodology
codingagentsystem/cas/CLAUDE.md
entry/task/rule/skill) ``` **Human CLI:** ```bash cas init # Initialize CAS in a project cas serve # Start MCP server cas config list # View configuration cas doctor # Run diagnostics cas update # Self-update ``` ## Project
nxtg-ai/forge-orchestrator/CLAUDE.md
cargo run -- dashboard --pty # TUI dashboard with native agent TUIs cargo run -- mcp # Start MCP server (stdio JSON-RPC 2.0) cargo run -- status # Task board + event log cargo
Open-Document-Alliance/PDF-Tools/CLAUDE.md
Toolkit extension. ## Project Overview PDF Toolkit is a Claude Desktop extension (MCPB) and MCP server that enables automated PDF form filling, bulk processing, and data extraction. Published in the official
shashankswe2020-ux/whoop-mcp/CLAUDE.md
format # Format with Prettier npm run typecheck # Type check (no emit) node dist/index.js # Run MCP server (production) ``` ## Project Structure ``` src/ ├── index.ts # Entry point — creates MCP server, authenticates, starts stdio ├── server.ts
toolprint/hypertool-mcp/CLAUDE.md
mcp**: a TypeScript MCP proxy server that routes requests between clients and multiple underlying MCP servers. ### 📋 Essential Reading **READ FIRST**: `.taskmaster/docs/prd.txt` - Contains complete project requirements, architecture, and technical specifications. ### 🎯 High
toolprint/hypertool-mcp/src/CLAUDE.md
HyperTool MCP Server - Usage Guide ## Overview HyperTool MCP is a unified MCP proxy server that can operate in two distinct modes: 1. **Persona Mode** - Using content packs with bundled MCP
raullenchai/vnsh/CLAUDE.md
browser UI, public-content host, cleanup - `cli/npm/`: Node CLI (`vn` and `vnsh`) - `mcp/`: MCP server for workspace create/read/update/open and legacy share/read - `extension/`: Chrome extension for previews and capture bundles - `scripts/smoke.mjs
TradersPost/pinescript-agents/CLAUDE.md
work here) - `templates/` - Ready-to-use templates - `examples/` - Reference implementations - `docs/` - Pine Script documentation + MCP server - `tools/` - Development utilities ## Essential Limits - Max 500 bars lookback - Max 500 plot outputs
Atomburstofficial/geiger/CLAUDE.md
counter for AI agents. One read-only command that inventories every AI agent, harness, MCP server, plugin, and AI extension on a machine and reports what each can touch. Published
AnastasiyaW/codex-claude-code-config/CLAUDE.md
order -> first release checklist. **When to use:** new Agent SDK app, custom orchestrator, new MCP server, new Cloudflare Worker with tool calls. **Not for:** improvement of an existing harness
awslabs/nx-plugin-for-aws/CLAUDE.md
example workspace using locally compiled versions: - Consult the Nx Plugin for AWS MCP server (`create-workspace-command`, `list-generators`, `generator-guide`, `general-guidance`) as the baseline for creating workspaces
SnakeO/claude-co-commands/CLAUDE.md
Code plugin (`snakeo-co-commands`) providing three collaboration skills that use OpenAI's Codex MCP server as a second opinion. No compiled code, no build steps, no tests — purely Markdown
bobmatnyc/claude-mpm/CLAUDE.md
pytest tests/path/test_x.py # single file ``` --- ## Architecture Overview ``` CLI (click) → Services → Agents → Claude Code subprocess ↓ Skills / MCP servers ``` **Deep dive**: See [`docs/developer/AGENT_ASSEMBLY_PIPELINE.md`](docs/developer/AGENT_ASSEMBLY_PIPELINE.md) for how agent definitions are produced, composed, and delivered
RaphaelRegnier/vibe-annotations/CLAUDE.md
agents via MCP. ## Repo structure ``` packages/ extension/ Chrome extension — vanilla JS, no build step server/ MCP server — npm package (vibe-annotations-server) website/ Marketing site & docs — Next.js, deployed on Vercel
AnkitClassicVision/Claude-Code-Deep-Research/CLAUDE.md
spawning autonomous agents for complex multi-step operations 5. **TodoWrite/TodoRead**: For tracking research progress ### MCP Server Tools: 1. **mcp__filesystem__**: File system operations (read, write, search files) 2. **mcp__puppeteer
byt3bl33d3r/figaro/patchright-cli/CLAUDE.md
design decisions:** - Every test is marked `@pytest.mark.integration` (registered in `pyproject.toml`) - No external HTTP server — tests use `data:text/html,...` URLs for deterministic DOM - Unix socket path length: `test_daemon_lifecycle.py` and `test_socket_server.py
ruanrongman/IntelliConnect/CLAUDE.md
featuring multi-agent architecture, voice capabilities (ASR/TTS/VAD), RAG knowledge bases, native knowledge graphs, MCP (Model Context Protocol) support, and complete IoT infrastructure with Thing Model abstraction. ## Technology Stack - **Backend**: Java
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.