Doctor

v1.0.0Developer Toolsstable

Lighthouse for MCP servers — CLI diagnostic and profiling tool that runs 8 checks against any MCP endpoint and produces a letter-graded report card with latency profiling, concurrency stress testing, schema validation, and CI/CD integration.

ai-agentscicdclideveloper-toolsdevops
Share:
465
Stars
0
Downloads
0
Weekly
0/5

What is Doctor?

Doctor is a Model Context Protocol (MCP) server that allows AI assistants like Claude, Cursor, and VS Code to lighthouse for mcp servers — cli diagnostic and profiling tool that runs 8 checks against any mcp endpoint and produces a letter-graded report card with latency profiling, concurrency stress testing, ...

Lighthouse for MCP servers — CLI diagnostic and profiling tool that runs 8 checks against any MCP endpoint and produces a letter-graded report card with latency profiling, concurrency stress testing, schema validation, and CI/CD integration.

This server falls under the Developer Tools category on MCPgee, the world's largest MCP server directory with 33,000+ servers.

Features

  • Lighthouse for MCP servers — CLI diagnostic and profiling to

Use Cases

MCP server diagnostics
Latency profiling and stress testing
CI/CD health checks
sisig-ai

Maintainer

LicenseMIT
Languagepython
Versionv1.0.0
UpdatedMay 6, 2026
Statushealthy
Maintenanceactive

Works with

ClaudeOpenAIwindowsmacoslinux

Installation

Manual Installation

npx doctor

Configuration

Configuration Details

Config File

claude_desktop_config.json

Performance

Response Metrics

Response Time< 200ms
ThroughputMedium

Resource Usage

Memory UsageLow
CPU UsageLow

How to Set Up and Use Doctor

Doctor is an MCP server and REST API that provides AI assistants with a crawled, embedded, and searchable knowledge base of any website. It uses crawl4ai for hierarchical web crawling, LangChain for text chunking, OpenAI embeddings via litellm for vector encoding, and DuckDB for vector storage and search. Once a site is indexed, AI assistants can search it semantically, retrieve pages, and navigate the site hierarchy — giving agents accurate, up-to-date documentation access rather than relying on stale training data.

Prerequisites

  • Docker and Docker Compose installed
  • Python 3.12 or higher (if running outside Docker)
  • An OpenAI API key for generating embeddings
  • An MCP client such as Cursor or VS Code to consume the MCP endpoint
1

Clone the repository

Clone the sisig-ai/doctor repository to your local machine.

git clone https://github.com/sisig-ai/doctor.git
cd doctor
2

Set your OpenAI API key

Export the OPENAI_API_KEY environment variable before starting the stack. This is used by litellm to generate embeddings for indexed content.

export OPENAI_API_KEY=your-openai-api-key
3

Start the full stack with Docker Compose

Run the docker compose command to bring up the FastAPI web server (port 9111), Redis task broker, and crawl worker together.

docker compose up
4

Crawl a website to build the knowledge base

With the stack running, use the /fetch_url endpoint to start indexing a site. Monitor progress via /job_progress.

# Start a crawl
curl -X POST http://localhost:9111/fetch_url \
  -H 'Content-Type: application/json' \
  -d '{"url": "https://docs.example.com"}'

# Check crawl progress
curl http://localhost:9111/job_progress
5

Configure your MCP client

Add the Doctor MCP endpoint to your editor's MCP configuration. The server uses SSE transport at the /mcp path.

Doctor Examples

Client configuration

Add Doctor to your Cursor or VS Code MCP config using the SSE transport. The server must be running locally via Docker Compose.

{
  "mcpServers": {
    "doctor": {
      "type": "sse",
      "url": "http://localhost:9111/mcp"
    }
  }
}

Prompts to try

Once Doctor has indexed a site and your MCP client is connected, try these prompts.

- "Search the indexed docs for how to configure authentication."
- "Show me the site map of the documentation you have indexed."
- "Find all pages related to the API reference and list their titles."
- "What does the getting-started guide say about environment variables?"
- "Navigate to the sibling pages of the deployment guide and summarise each."

Troubleshooting Doctor

Docker Compose fails to start with missing environment variable errors

Ensure OPENAI_API_KEY is exported in your shell before running 'docker compose up', or add it to a .env file in the project root that Docker Compose will pick up automatically.

Crawl jobs are submitted but never complete

Check that the Redis container is healthy with 'docker compose ps'. The crawl worker communicates with the FastAPI server via Redis; if Redis is not running, jobs will queue indefinitely.

MCP client cannot connect to http://localhost:9111/mcp

Verify the Docker stack is running with 'docker compose ps' and that port 9111 is not blocked by a firewall or already in use. Check container logs with 'docker compose logs' for startup errors.

Frequently Asked Questions about Doctor

What is Doctor?

Doctor is a Model Context Protocol (MCP) server that lighthouse for mcp servers — cli diagnostic and profiling tool that runs 8 checks against any mcp endpoint and produces a letter-graded report card with latency profiling, concurrency stress testing, schema validation, and ci/cd integration. It connects AI assistants to external tools and data sources through a standardized interface.

How do I install Doctor?

Follow the installation instructions on the Doctor GitHub repository. Clone the repo, install dependencies, and add the server config to your AI client.

Which AI clients work with Doctor?

Doctor works with all major MCP-compatible AI clients including Claude Desktop, Claude Code, Cursor, VS Code (GitHub Copilot), Windsurf, and Cline.

Is Doctor free to use?

Yes, Doctor is open source and available under the MIT license. You can use it freely in both personal and commercial projects.

Browse More Developer Tools MCP Servers

Explore all developer tools servers available in the MCPgee directory. Each server includes setup guides for Claude, Cursor, and VS Code.

Quick Config Preview

{ "mcpServers": { "doctor": { "command": "npx", "args": ["-y", "doctor"] } } }

Add this to your claude_desktop_config.json or .cursor/mcp.json

Read the full setup guide →

Ready to use Doctor?

Browse our complete directory of 33,000+ MCP servers, read setup guides for your editor, and start building with the Model Context Protocol.

33,000+ ServersFree & Open SourceStep-by-Step Guides