RAG Documentation Server

Example Use Cases

• 1

  Search and retrieve relevant documentation snippets and source links via semantic vector search to ground responses.

• 2

  Index and make local code repositories searchable, including intelligent chunking and change-watching for up-to-date context.

• 3

  Ingest and manage large collections of web documentation using queue-based processing, deduplication, and embedding fallbacks.

Description

This server provides a suite of MCP tools for semantic search, source management, URL extraction, queue-based ingestion, and local repository indexing. It generates embeddings (using Ollama by default, with OpenAI as a fallback), chunks content intelligently, and stores vectors for fast retrieval. The project includes a web UI, Docker Compose setup, and asynchronous repository indexing with progress monitoring to avoid timeouts when processing large codebases.