mcp-registry/mcp-server-deep-research

    ==================
      
       /// MCP ///
      /// MCP ///
        
    ==================
        
    [server:online]
    [protocol:ready]

mcp-server-deep-research

by reading-plus-ai

MCP Server for Deep Research – a Python-based MCP server that turns a research question into a multi-step workflow (question elaboration → sub-question generation → web search → content analysis → report generation) and returns a fully cited research report.

150
15
Open Source
GitHub

01

Question Elaboration

Expands and clarifies a research question, identifies key terms and concepts, and defines the scope and parameters for the study.

02

Subquestion Generation

Produces focused subquestions that address different aspects of the main topic, ensuring comprehensive coverage and a structured research plan.

03

Web Search Integration

Uses Claude's built-in web search to run targeted searches for each subquestion, surface authoritative sources, and gather diverse perspectives.

04

Content Analysis

Evaluates the quality and relevance of gathered information, synthesizes findings across sources, and generates proper citations.

05

Report Generation

Creates well-structured, comprehensive research reports that cite all sources and present balanced, evidence-based conclusions.

06

deep-research

A prompt template specifically designed to guide comprehensive, structured research workflows.

Installation

1. Prerequisites
• Python 3.10 or newer
• Git
2. Clone the repository
git clone https://github.com/reading-plus-ai/mcp-server-deep-research.git cd mcp-server-deep-research
3. Create and activate a virtual environment (recommended)
python -m venv venv # macOS/Linux source venv/bin/activate # Windows venv\Scripts\activate
4. Install Python dependencies
pip install -r requirements.txt
5. Configuration
• If a .env.example file is present, copy it and supply your own values:
• Set any required API keys or DB connection strings used by the MCP server.
cp .env.example .env
6. Run the development server
# FastAPI-style (most common for recent MCP servers) uvicorn app.main:app --host 0.0.0.0 --port 8000 --reload # Flask-style alternative python app/main.py
7. Verify
Open http://localhost:8000/ (or /docs) to confirm the server is up.
8. Production deployment
• Use Gunicorn/Uvicorn workers behind a reverse proxy such as Nginx.
• Enable HTTPS and set proper environment variables for production.

Documentation

License: MIT License
Updated 7/30/2025