Tea Rags is a high-performance MCP server designed for deep, intelligent codebase analysis, capable of handling enterprise-scale projects with millions of lines of code. It excels with fast local indexing and offers advanced features like semantic code search, hybrid retrieval, and filtering by git-aware metadata such as authorship, code churn, and task IDs. Built on Qdrant, Tea Rags prioritizes privacy by supporting local deployments with Ollama, while also offering compatibility with major cloud providers like OpenAI, Cohere, and Voyage AI. It features incremental indexing for efficient updates, AST-aware chunking for high accuracy across various languages, and comprehensive performance tuning capabilities, making it a robust and flexible solution for modern software development.
Key Features
015 GitHub stars
02Built for scale with fast local indexing for enterprise codebases (millions of LOC).
03Git-aware search with filters by author, code age, churn, and task IDs.
04Semantic and hybrid code search with AST-aware chunking for accuracy.
05Privacy-first operation with local Ollama support and incremental indexing.
06Highly configurable for performance tuning across various hardware and embedding providers.
