Provides LLMs with semantic search capabilities over Markdown documentation stored in S3, leveraging 100% local embeddings to eliminate API costs.
This tool acts as an MCP server designed to empower large language models with advanced semantic search over extensive documentation stored in S3 buckets. It enables AI clients to intelligently query and retrieve relevant information from Markdown files, utilizing a local RAG (Retrieval Augmented Generation) system powered by Ollama embeddings. The server features smart, incremental synchronization with S3, fast vector-based search, and exposes specific MCP tools for querying documentation and refreshing its index, making it ideal for creating AI-powered knowledge bases without external API dependencies.