Light Research

A lightweight MCP server for LLM orchestration with DuckDuckGo/GitHub Code search and content extraction
  • typescript

7

GitHub Stars

typescript

Language

6 months ago

First Indexed

2 months ago

Catalog Refreshed

Documentation & install

Readme and setup notes from the catalogue, plus a client-ready config you can copy for your MCP host.

Installation

Add the following to your MCP client configuration file.

Configuration

View docs
{
  "mcpServers": {
    "code-hex-light-research-mcp": {
      "command": "llmresearcher",
      "args": [
        "--mcp"
      ],
      "env": {
        "TIMEOUT": "30000",
        "USER_AGENT": "Mozilla/5.0 (compatible; LLMResearcher/1.0)",
        "MAX_RESULTS": "10",
        "MAX_RETRIES": "3",
        "CACHE_ENABLED": "true",
        "RATE_LIMIT_DELAY": "1000"
      }
    }
  }
}

You are using a lightweight MCP server that orchestrates LLM workflows for web search and content extraction. It exposes tools that let your model search the web, fetch code examples, and extract clean, LLM-friendly content from pages, all in a fast, privacy-conscious package.

How to use

Start by running the MCP server in stdio mode to provide search and extraction tools to your LLM client. The server exposes these capabilities: github_code_search, duckduckgo_web_search, and extract_content. You can use an MCP-enabled client to issue tasks that combine web search with content extraction, returning clean, Markdown-formatted results suitable for LLM consumption.

How to install

Prerequisites you need before installing and running:

  • Node.js 20.0.0 or higher
  • No local Chrome installation required (Playwright uses its bundled Chromium)

Install and set up the project step by step.

# Clone or download the project
cd light-research-mcp

# Install dependencies (using pnpm)
pnpm install

# Build the project
pnpm build

# Install Playwright browsers
pnpm install-browsers

# Optional: Link globally for system-wide access
pnpm link --global

Start the MCP server (stdio transport)

To run the MCP server so your LLM client can connect over stdio, start the executable with the MCP flag.

llmresearcher --mcp

Add and manage the MCP server in Claude Code (example client)

If you use Claude Code, you can register the MCP server so your agents can access it. The following demonstrates adding the local MCP server and listing or checking its status.

claude mcp add light-research-mcp /path/to/light-research-mcp/dist/bin/llmresearcher.js --mcp

claude mcp add light-research-mcp -s project /path/to/light-research-mcp/dist/bin/llmresearcher.js --mcp

claude mcp list

claude mcp get light-research-mcp

MCP Tool Usage Concepts

Once configured, you can call the available MCP tools from your client to perform combined search and extraction tasks. The core capabilities are the following tools:

  • github_code_search: Search GitHub repositories for code examples and implementation patterns
  • duckduckgo_web_search: Search the web with DuckDuckGo and retrieve results
  • extract_content: Extract detailed content from URLs using a headless browser and readability processing

Available tools

github_code_search

Search GitHub repositories for code examples and implementation patterns using MCP tooling.

duckduckgo_web_search

Perform web searches via DuckDuckGo HTML endpoint and retrieve results optimized for LLM use.

extract_content

Navigate to a URL with a headless browser, render JS, then extract main content using Readability and sanitize to Markdown.

Built by
VeilStrat
AI signals for GTM teams
© 2026 VeilStrat. All rights reserved.All systems operational