MCP Local RAG

Fit and risk

Setup

{"mcpServers":{"local-rag":{"command":"npx","args":["-y","mcp-local-rag"],"env":{"BASE_DIR":"/path/to/your/documents"}}}}

{
  "servers": {
    "local-rag": {
      "command": "npx",
      "args": [
        "-y",
        "mcp-local-rag"
      ],
      "env": {
        "BASE_DIR": "/path/to/your/documents"
      }
    }
  }
}

claude mcp add local-rag -e BASE_DIR=/path/to/your/documents -- npx -y mcp-local-rag

Check that it works

After restarting your MCP client, the 9 tools (query_documents, ingest_file, etc.) should appear in the tool list; then ask a question about an ingested document to confirm search returns results.

Troubleshooting

1. Verify that documents were ingested first, then use list_files or status to inspect indexing state
2. If the first model download fails, check the network connection or proxy configuration
3. For file-size errors, inspect or adjust MAX_FILE_SIZE
4. For path-restriction errors, confirm the path is inside BASE_DIR, BASE_DIRS, or a CLI-configured root
5. Check MCP configuration syntax and fully restart the client; run npx mcp-local-rag directly to verify startup

Things to try

Once connected, you can ask your AI assistant things like:

• Ingest the document at /Users/me/docs/api-spec.pdfCopy
• What does the API documentation say about authentication?Copy
• List all files in the configured base directories and their ingested statusCopy
• Sync the index with /Users/me/docs and tell me when it's finishedCopy

Tools 9

No matching tools

Use cases

Supported clients

Listed from the project's documentation, not tested by this site.

Overview

Similar servers