{
  "markdown": "# Meta-Prompt MCP\n\n> Instant access to Google and Anthropic's official prompting guides within your LLM workflow—optimized for crafting high-quality meta-prompts and system prompts.\n\n[![Python 3.10+](https://img.shields.io/badge/python-3.10+-blue.svg)](https://www.python.org/downloads/)\n[![License: MIT](https://img.shields.io/badge/License-MIT-green.svg)](LICENSE)\n\n---\n\n## What It Does\n\nMeta-Prompt MCP is an MCP server that surfaces official prompting best practices from Google and Anthropic directly within your LLM workflow. Instead of searching documentation or guessing how to instruct an LLM, you can query expert guides on-demand.\n\nThis is especially valuable when crafting meta-prompts (system prompts for agents) or optimizing your own prompting strategies. By grounding your prompts in proven methodology, you'll generate more effective, well-structured outputs.\n\n### How It Works\n\nYour MCP host (Claude Desktop, Cursor, etc.) communicates with the server via stdio. When you ask for a guide, the server retrieves it from bundled markdown files—no API calls, no latency.\n\n```\n┌────────────────────────┐          ┌──────────────────────────┐\n│   Your LLM Host        │◄─stdio──►│  Meta-Prompt MCP Server  │\n│  (Claude Desktop, etc) │          │                          │\n└────────────────────────┘          │  • get_google_guide      │\n                                    │  • get_anthropic_guide   │\n                                    │                          │\n                                    │  Data layer:             │\n                                    │  ./data/                 │\n                                    │  ├── google_*.md         │\n                                    │  └── anthropic_*.md      │\n                                    └──────────────────────────┘\n```\n\n### Key Features\n\n- **`get_google_guide`** — Retrieves Google's comprehensive prompting guide covering techniques, best practices, and LLM configuration\n- **`get_anthropic_guide`** — Retrieves Anthropic's guide on chain-of-thought, multishot prompting, and extended thinking\n- **Zero dependencies** — Runs entirely offline with bundled guides; no API keys or network calls required\n\n---\n\n## Validation\n\nWe ran a benchmark comparing prompts generated with and without guide access across 5 diverse tasks. An independent judge LLM scored each on Clarity, Specificity, Structure, Effectiveness, and Overall quality (1–10 scale).\n\n**[See full results →](benchmarks/results.md)**\n\nTo reproduce the benchmark:\n\n```bash\nexport OPENROUTER_API_KEY=sk-or-...\nmake benchmark\n```\n\n---\n\n## Quick Start\n\n### 1. Install\n\n```bash\n# Via uvx (recommended — run without installing globally)\nuvx meta-prompt-mcp\n\n# Or install via pip\npip install meta-prompt-mcp\n```\n\nThe package ships with bundled markdown guides — no API keys or setup needed.\n\n### 2. Configure Your MCP Host\n\n#### Claude Desktop\n\nAdd to your `claude_desktop_config.json`:\n\n```json\n{\n  \"mcpServers\": {\n    \"meta-prompt-mcp\": {\n      \"command\": \"uvx\",\n      \"args\": [\"meta-prompt-mcp\"]\n    }\n  }\n}\n```\n\n#### Cursor\n\nAdd to your MCP settings:\n\n```json\n{\n  \"mcpServers\": {\n    \"meta-prompt-mcp\": {\n      \"command\": \"uvx\",\n      \"args\": [\"meta-prompt-mcp\"]\n    }\n  }\n}\n```\n\n#### Claude Code\n\nRun the following command in your terminal:\n\n```bash\nclaude mcp add meta-prompt-mcp -- uvx meta-prompt-mcp\n```\n\n---\n\n## Usage\n\nQuery the guides while crafting your prompts. Examples:\n\n- *\"I'm building a code reviewer agent. Reference the Google guide to help me write a better system prompt.\"*\n- *\"Based on the Anthropic guide, suggest improvements to this prompt for better reasoning.\"*\n- *\"What technique from the guides would work best for this task?\"*\n\nThe LLM reads the guides and uses that knowledge to give you more informed suggestions and feedback on your prompts.\n\n---\n\n## Development\n\n```bash\n# Clone the repo\ngit clone https://github.com/kapillamba4/meta-prompt-mcp.git\ncd meta-prompt-mcp\n\n# Install in dev mode\nmake dev\n\n# Run the server\nmake run\n```\n\n### Available Commands\n\n```bash\nmake dev        # Install in editable mode with dev dependencies\nmake run        # Start the MCP server locally\nmake lint       # Check code quality\nmake format     # Auto-format code with Black & isort\nmake test       # Run test suite\nmake benchmark  # Run prompt quality benchmark (requires OPENROUTER_API_KEY)\nmake build      # Build distribution packages\nmake publish    # Publish to PyPI\n```\n\n---\n\n## Project Structure\n\n```\nmeta-prompt-mcp/\n├── pyproject.toml                 # Dependencies and package config\n├── Makefile                       # Development commands\n├── .env.example                   # Template for OPENROUTER_API_KEY\n│\n├── benchmarks/\n│   ├── benchmark.py               # Prompt quality evaluation script\n│   └── results.md                 # Benchmark results\n│\n└── src/meta_prompt_mcp/\n    ├── __init__.py\n    ├── __main__.py                # Entry point (python -m)\n    ├── server.py                  # FastMCP server & tool definitions\n    └── data/\n        ├── google_prompting_guide.md\n        └── anthropic_prompting_guide.md\n```\n\n---\n\n## Contributing\n\nIssues, feature requests, and contributions welcome. Please open an issue on [GitHub](https://github.com/kapillamba4/meta-prompt-mcp/issues).\n\n## License\n\nMIT\n",
  "bytes": 5286,
  "sha": "319f8f56f794b20c061417de5f76d7cbc2b30a8558d6c04ade7a1d5cfa4a50f9",
  "repo_slug": "kapillamba4/meta-prompt-mcp",
  "fonte": "repo",
  "truncated": false,
  "api": "https://agentalog.com/api/listings/mcp_io_github_kapillamba4_meta_prompt_mcp_1c2409ea/readme"
}