Local RAG Search - Agent Skill

Name: Local RAG Search: AI Agent Skill
Rating: 9 (1 reviews)
Author: nkapila6

An Agent Skill that teaches Claude how to effectively use the mcp-local-rag MCP server for intelligent web searches with semantic similarity ranking.

What This Skill Does

This skill enables agents to:

Choose the right search tool based on the task (DuckDuckGo, Google, or multi-engine deep research)
Formulate effective queries using natural language
Tune parameters for different use cases (quick answers vs comprehensive research)
Perform deep research across multiple search engines and topics
Respect privacy by defaulting to DuckDuckGo

Prerequisites

This skill requires the mcp-local-rag MCP server to be installed and configured in your MCP client.

Install mcp-local-rag

Add to your MCP configuration:

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

Or use Docker:

{
  "mcpServers": {
    "mcp-local-rag": {
      "command": "docker",
      "args": [
        "run", "--rm", "-i", "--init",
        "-e", "DOCKER_CONTAINER=true",
        "ghcr.io/nkapila6/mcp-local-rag:v1.0.2"
      ]
    }
  }
}

Installation

Claude Desktop

Navigate to Settings → Skills
Click Add Skill → Add from folder
Select this skill folder (local-rag-search/)

Usage

Once both the MCP server and skill are loaded, simply ask Claude to search for information:

"Search the web for the latest Python 3.13 features"
"Do deep research on sustainable energy solutions"
"Find technical documentation about Docker optimization"

Claude will automatically apply the skill's best practices to use the appropriate tools effectively.

Features

✅ Smart tool selection - Automatically chooses DuckDuckGo, Google, or deep research based on query
✅ Privacy-first - Defaults to DuckDuckGo for general searches
✅ Multi-engine research - Supports 9+ search backends for comprehensive coverage
✅ Semantic ranking - Uses RAG-like similarity scoring for most relevant results
✅ No external APIs - All processing runs locally with embedded models

Supported Search Backends

DuckDuckGo (privacy-focused)
Google (comprehensive)
Bing, Brave, Yahoo, Yandex
Wikipedia (factual/encyclopedia)
Mojeek, Grokipedia

License

MIT - Same as the parent mcp-local-rag project

Frequently Asked Questions

What is Local RAG Search?

Local RAG Search is a free, open-source AI agent skill. Efficiently perform web searches using the mcp-local-rag server with semantic similarity ranking. Use this skill when you need to search the web for current information, research topics across multiple sources, or gather context from the internet without using external APIs. This skill teaches effective use of RAG-based web search with DuckDuckGo, Google, and multi-engine deep research capabilities.

How do I install Local RAG Search?

Install Local RAG Search with a single command:

npx mdskills install nkapila6/local-rag-search

This downloads the skill files into your project and your AI agent picks them up automatically.

What platforms support Local RAG Search?

Local RAG Search works with Claude Code, Claude Desktop, Cursor, Vscode Copilot, Windsurf, Continue Dev, Gemini Cli, Amp, Roo Code, Goose. Skills use the open SKILL.md format which is compatible with any AI coding agent that reads markdown instructions.

Local RAG Search