DeepResearch

Provides a structured MCP workflow to run deep research sessions with automated subquestions, layered analysis, and publication-quality reports.

other

0

GitHub Stars

other

Language

4 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": {
    "lihongwen-deepresearch-mcpserver": {
      "command": "uv",
      "args": [
        "--directory",
        "/path/to/your/deepresearch-mcpserver",
        "run",
        "mcp-server-deep-research"
      ]
    }
  }
}

You can run the Deep Research MCP Server locally and access a guided, multi-layer research workflow from your MCP client. It orchestrates layered information gathering, critical analysis, and publication‑quality reports, helping you produce well-structured research outputs with clear evidence and citations.

How to use

Start a research session by invoking the start_deep_research tool through your MCP client. You provide a research question, and the server guides you through a structured workflow: clarifying concepts, decomposing into subquestions, gathering information in layered stages, performing critical analysis, and generating a comprehensive report. The workflow adapts depth and methodology to the complexity of your question, with an executive summary, methodology, findings, synthesis, and actionable recommendations.

How to install

Prerequisites you need before installing the MCP server are in place to ensure compatibility and smooth operation.

Install the required runtime and tooling:

Install Python 3.10 or higher
Install the uv runtime (the MCP runner) for local servers

Step-by-step commands you can follow to set up the local MCP server environment exactly as described:

Additional details

Configuration in your Claude Desktop client is required to connect to the local MCP server instance. The configuration maps a labeled MCP server to the local runtime, so Claude Desktop can start, monitor, and receive the final reports from the MCP workflow.

{
  "mcpServers": {
    "mcp-server-deep-research": {
      "command": "uv",
      "args": [
        "--directory",
        "/path/to/your/deepresearch-mcpserver",
        "run",
        "mcp-server-deep-research"
      ]
    }
  }
}

Available tools

start_deep_research

Direct tool to initiate a comprehensive deep research session with guided workflow and structured outputs.