Mcp Local Rag

What is Mcp Local Rag about?

Mcp Local Rag enables language models to perform live web searches, extract relevant content, and incorporate that context into their responses without relying on external APIs. It runs entirely on the user’s machine, making it ideal for privacy‑focused or offline workflows.

How to use Mcp Local Rag?

1. Add to MCP configuration – Insert a server entry for mcp-local-rag in the MCP client settings, either using the uvx command (Python) or Docker.
2. Trigger from an LLM – When a model detects a need for up‑to‑date information, it calls the mcp-local-rag tool.
3. Receive markdown context – The tool returns the extracted HTML content as markdown, which the model then includes in its final answer.

Key features of Mcp Local Rag

• Searches DuckDuckGo and fetches the top 10 results.
• Generates embeddings using Google’s MediaPipe Text Embedder.
• Computes similarity between query and results to rank entries.
• Extracts and converts HTML content to markdown for easy consumption.
• Works with any MCP‑compatible client that supports tool calling (e.g., Claude Desktop, Cursor, Goose).
• No external API keys required; runs fully locally.

Use cases of Mcp Local Rag

• Retrieving recent news or product releases that the base model is unaware of.
• Supplementing LLM responses with factual data from the web in real time.
• Privacy‑preserving search where data never leaves the local environment.
• Prototyping RAG pipelines without setting up cloud services.

FAQ from the Mcp Local Rag

Q: Do I need an API key? A: No. The tool uses DuckDuckGo for search and MediaPipe for embeddings, both of which are free and do not require authentication.

Q: Which programming language is required? A: The server is Python‑based. Installation can be done via uvx (recommended) or Docker.

Q: Can I customize the number of search results or the embedding model? A: Yes. The source code exposes configuration variables for result count and the embedding pipeline.

Q: Is it compatible with all LLMs? A: It works with any MCP client that supports tool calling, which includes most major chat interfaces that allow custom tools.

Q: How does it handle rate limits or bans? A: Using DuckDuckGo mitigates strict rate limiting, but heavy usage should respect typical search etiquette.

mcp-local-rag

"primitive" RAG-like web search model context protocol (MCP) server that runs locally. ✨ no APIs ✨

Installation

Locate your MCP config path here or check your MCP client settings.

Run Directly via uvx

This is the easiest and quickest method. You need to install uv for this to work. Add this to your MCP server configuration:

{
  "mcpServers": {
    "mcp-local-rag":{
      "command": "uvx",
        "args": [
          "--python=3.10",
          "--from",
          "git+https://github.com/nkapila6/mcp-local-rag",
          "mcp-local-rag"
        ]
      }
  }
}

Using Docker (recommended)

Ensure you have Docker installed. Add this to your MCP server configuration:

{
  "mcpServers": {
    "mcp-local-rag": {
      "command": "docker",
      "args": [
        "run",
        "--rm",
        "-i",
        "--init",
        "-e",
        "DOCKER_CONTAINER=true",
        "ghcr.io/nkapila6/mcp-local-rag:latest"
      ]
    }
  }
}

Security audits

MseeP does security audits on every MCP server, you can see the security audit of this MCP server by clicking here.

MCP Clients

The MCP server should work with any MCP client that supports tool calling. Has been tested on the below clients.

• Claude Desktop
• Cursor
• Goose
• Others? You try!

Examples on Claude Desktop

When an LLM (like Claude) is asked a question requiring recent web information, it will trigger mcp-local-rag.

When asked to fetch/lookup/search the web, the model prompts you to use MCP server for the chat.

In the example, have asked it about Google's latest Gemma models released yesterday. This is new info that Claude is not aware about.

Result

mcp-local-rag performs a live web search, extracts context, and sends it back to the model—giving it fresh knowledge:

Contributing

Have ideas or want to improve this project? Issues and pull requests are welcome!

License

This project is licensed under the MIT License.