rag-server-mcp
If you are the rightful owner of rag-server-mcp and would like to certify it and/or have it hosted online, please leave a comment on the right or send an email to henry@mcphub.com.
mcp-rag-server is a Model Context Protocol server that enables Retrieval Augmented Generation capabilities for connected LLMs.
The MCP RAG Server is a specialized server designed to enhance the capabilities of connected Language Learning Models (LLMs) by providing Retrieval Augmented Generation (RAG) functionalities. It achieves this by indexing documents from a project and supplying relevant context to LLMs, thereby improving their response accuracy and relevance. Built using Google Genkit, ChromaDB, and Ollama, the server offers seamless integration within the Model Context Protocol ecosystem. It emphasizes local control and privacy by utilizing local models and vector stores. The server is designed to be extensible, allowing for future enhancements and integrations. It supports automatic indexing of various file types and uses hierarchical chunking for efficient data processing. The server is Dockerized for easy setup and deployment, making it accessible for developers looking to integrate RAG capabilities into their projects.
Features
- Automatic Indexing: Scans and indexes project files automatically on startup.
- Supported File Types: Handles .txt, .md, code files, .json, .jsonl, .csv.
- Hierarchical Chunking: Separates text and code blocks in Markdown files.
- Vector Storage: Utilizes ChromaDB for persistent vector storage.
- Local Embeddings: Uses Ollama for local embedding generation.
