MCPO - MCP Pollinations Proxy

# CLAUDE.md This file provides guidance to Claude Code (claude.ai/code) when working with code in this repository. ## Project Overview This is a Docker-containerized MCP (Model Context Protocol) proxy setup that combines: - **mcpo**: A Python CLI tool that creates an OpenAPI proxy for MCP servers, making them accessible via standard REST endpoints - **Pollinations MCP**: A Node.js package that provides AI image, text, and audio generation capabilities through the Pollinations API The architecture creates a bridge between MCP servers and applications expecting OpenAPI/REST interfaces. ## Development Commands ### Docker Operations ```bash # Build the container docker-compose build # Run the service docker-compose up # Run in detached mode docker-compose up -d # View logs docker-compose logs # Stop the service docker-compose down ``` ### Service Access - The service runs on port 7777 - OpenAPI documentation available at: http://localhost:7777/docs - The proxy exposes Pollinations MCP server functionality via REST endpoints ## Architecture Details ### Container Setup - **Base**: Node.js 18 Alpine Linux - **Python**: Installed alongside Node.js for mcpo CLI - **Port**: 7777 exposed for HTTP access - **Host**: Configured to bind to 0.0.0.0 for external access ### Service Flow 1. Container starts with `mcpo` CLI tool 2. `mcpo` proxies the `pollinations-model-context-protocol` MCP server 3. MCP server capabilities become available via OpenAPI endpoints 4. External applications can use standard HTTP/REST calls instead of MCP protocol ### Key Components - **mcpo CLI**: Converts MCP servers to OpenAPI REST endpoints - **Pollinations MCP**: Provides AI generation tools (images, text, audio) - **Docker**: Containerizes the entire stack for easy deployment ## Package Management - Node.js packages installed globally via npm - Python packages installed system-wide via pip - No package.json or requirements.txt files in this repository