Docs RAG icon

Docs RAG

Createdkazuph

Enables querying of local documents using retrieval-augmented generation (RAG) with LLMs.

About

This tool is a TypeScript-based server that implements a RAG (Retrieval-Augmented Generation) system for documents stored in a local directory. It allows users to query local Git repositories and text files using LLMs, leveraging the content of these documents as context for generating responses. The server provides tools for adding documents from Git repositories or direct text files, listing available documents, and performing RAG queries, integrating seamlessly with platforms like Claude Desktop to enhance conversational AI with local knowledge.

Key Features

  • Lists and accesses documents via `docs://` URIs.
  • Provides `list_documents` tool to list available documents.
  • 1 GitHub stars
  • Supports Git repositories and text files as document sources.
  • Offers `rag_query` tool to query documents using RAG.
  • Includes `add_git_repository` and `add_text_file` tools for adding new documents.

Use Cases

  • Extracting information from local text files using natural language queries.
  • Querying internal documentation stored as Git repositories.
  • Augmenting LLM responses with context from local knowledge bases.
Craft Better Prompts with AnyPrompt