RAG Retriever

What it does

RAG Retriever connects your AI assistant to a custom knowledge base by crawling web pages and local files, indexing them into a vector store, and allowing semantic search based on natural language queries.

Tools

• rag-list-collections: Discover available vector store collections and metadata.
• rag-search-knowledge: Perform semantic search across specific or all collections with relevance scoring.
• rag-index-website: Crawl and index website content with configurable depth.
• rag-audit-collections: Review collection health and get maintenance recommendations.
• rag-assess-quality: Systematically evaluate content quality and reliability.
• rag-manage-collections: Perform administrative cleanup and deletion via CLI commands.
• rag-ingest-content: Guide advanced ingestion for local files, images, and enterprise systems.

Installation

Add the following to your claude_desktop_config.json:

{
  "mcpServers": {
    "rag-retriever": {
      "command": "python3",
      "args": ["/path/to/rag-retriever/mcp/server.py"]
    }
  }
}

Supported hosts

• Claude Desktop
• Claude Code
• Cursor