Deepresearch

🚀 Deep Research MCP Project Documentation

Deep Research MCP (Deep Research Model Context Protocol) is an intelligent research platform built on the Model Context Protocol, supporting integration with tools like Claude Desktop. This system offers powerful research capabilities by invoking OpenAI and Firecrawl APIs.

🚀 Quick Start

Deep Research MCP is an intelligent research platform that provides powerful research capabilities. Here's how to get started.

✨ Features

• Built on the Model Context Protocol.
• Supports integration with tools like Claude Desktop.
• Offers powerful research functions by calling OpenAI and Firecrawl APIs.

📦 Installation

Clone the Repository

git clone https://github.com/yourusername/deep-research-mcp.git
cd deep-research-mcp
npm install

Start the Service

node dist/index.js

📚 Documentation

Configuration Instructions

Research Initialization Tool

interface ResearchInitialization {
  query: string;
  depth: number;
}

Research Step Execution Tool

interface ResearchStepExecution {
  sessionId: string;
}

Final Report Generation Tool

interface ReportGeneration {
  sessionId: string;
  timeout?: number;
}

Client Code Examples

import { Client } from "@modelcontextprotocol/sdk/client/index.js";
import { StdioClientTransport } from "@modelcontextprotocol/sdk/client/stdio.js";

async function main() {
  // Connect to the server
  const transport = new StdioClientTransport({
    command: "node",
    args: ["dist/index.js"]
  });

  const client = new Client({ name: "deep-research-client", version: "1.0.0" });
  await client.connect(transport);

  // Initialize the research
  const initResult = await client.callTool({
    name: "initialize-research",
    arguments: {
      query: "The impact of artificial intelligence on healthcare",
      depth: 3
    }
  });
  
  // Parse the response to get the sessionId
  const { sessionId } = JSON.parse(initResult.content[0].text);
  
  // Execute steps until completion
  let currentDepth = 0;
  while (currentDepth < 3) {
    const stepResult = await client.callTool({
      name: "execute-research-step",
      arguments: { sessionId }
    });
    
    const stepInfo = JSON.parse(stepResult.content[0].text);
    currentDepth = stepInfo.currentDepth;
    
    console.log(`Completed step ${stepInfo.currentDepth}/${stepInfo.maxDepth}`);
  }
  
  // Generate the final report with a timeout setting
  const report = await client.callTool({
    name: "generate-report",
    arguments: { 
      sessionId,
      timeout: 180000 // Set a 3-minute timeout
    }
  });
  
  console.log("Final report:");
  console.log(report.content[0].text);
}

main().catch(console.error);

Troubleshooting

Common Issues

• Exceeding Token Limit: For very large research topics, you may encounter OpenAI token limit errors. Solutions:

  • Reduce the research depth.
  • Use more specific queries.
  • Break down complex topics into smaller sub - topics.

• Timeout Error: For complex research, the process may time out. Solutions:

  • Increase the timeout parameter for tool calls.
  • Use the complete-research tool and set a longer timeout.
  • Process research content in stages.

• API Rate Limit: If you encounter rate limit errors from OpenAI or Firecrawl:

  • Add delays between research steps.
  • Use API keys with higher rate limits.
  • Process requests in batches.

📄 License

This project is licensed under the MIT License. Please refer to the LICENSE file for details.

💡 Contribution Guidelines

Contributions are welcome! Please refer to CONTRIBUTING.md to learn how to contribute to the project.