ZoomEye MCP Server

A Model Context Protocol (MCP) server that provides network asset information based on query conditions. This server allows Large Language Models (LLMs) to obtain network asset information by querying ZoomEye using dorks and other search parameters.

This MCP server integrates with AI assistants and development environments like Claude Desktop, Cursor, Windsurf, Cline, Continue, and Zed, enabling them to search for and analyze internet-connected devices, services, and vulnerabilities through natural language interactions.

Features

• Query ZoomEye for network asset information using dorks
• Caching mechanism to improve performance and reduce API calls
• Automatic retry mechanism for failed API requests
• Comprehensive error handling and logging

Available Tools

• zoomeye_search - Get network asset information based on query conditions.
  • Required parameters:
    • qbase64 (string): Base64 encoded query string for ZoomEye search
  • Optional parameters:
    • page (integer): View asset page number, default is 1
    • pagesize (integer): Number of records per page, default is 10, maximum is 1000
    • fields (string): The fields to return, separated by commas
    • sub_type (string): Data type, supports v4, v6, and web. Default is v4
    • facets (string): Statistical items, separated by commas if there are multiple
    • ignore_cache (boolean): Whether to ignore the cache

Usage Guide

Basic Usage

Once the server is running, you can interact with it through your AI assistant or development environment. Here's how to use it:

1. Start the server using one of the installation methods above
2. Configure your AI assistant (Claude Desktop, Cursor, Windsurf, Cline, Continue, Zed, etc.) to use the server
3. Query network information using natural language

Search Syntax Guide

• Search Scope covers devices (IPv4, IPv6) and websites (domains).
• When entering a search string, the system will match keywords in "global" mode, including content from various protocols such as HTTP, SSH, FTP, etc. (e.g., HTTP/HTTPS protocol headers, body, SSL, title, and other protocol banners).
• Search strings are case-insensitive and will be segmented for matching (the search results page provides a " segmentation" test feature). When using == for search, it enforces exact case-sensitive matching with strict syntax.
• Please use quotes for search strings (e.g., "Cisco System" or 'Cisco System'). If the search string contains quotes, use the escape character, e.g.,"a"b". If the search string contains parentheses, use the escape character, e.g., portinfo().

You can see more detailed search syntax rules in prompts.py.

For more information on the ZoomEye Search API, refer to the ZoomEye API v2 documentation.

Getting Started

Prerequisites

1. ZoomEye API Key

   • Register for an account at ZoomEye
   • Obtain your API key from your account settings
   • The API key will be used to authenticate your requests to the ZoomEye API

2. Python Environment

   • Python 3.10 or higher is required
   • Alternatively, you can use Docker to run the server without installing Python

Installation

Using PIP

Alternatively, you can install mcp-server-zoomeye via pip:

pip install mcp-server-zoomeye

After installation, you can run it as a script using the following command:

python -m mcp_server_zoomeye

Using Docker

You can also run the ZoomEye MCP server using Docker:

Pull from Docker Hub

# Pull the latest image
docker pull zoomeyeteam/mcp-server-zoomeye:latest

# Run the container with your API key
docker run -i --rm -e ZOOMEYE_API_KEY=your_api_key_here zoomeyeteam/mcp-server-zoomeye:latest

Note: We provide multi-architecture Docker images that support linux/amd64 and linux/arm64 platforms and can run on Intel/AMD and ARM (such as Apple Silicon) processors.

Build from Source

Alternatively, you can build the Docker image from source:

# Clone the repository
git clone https://github.com/zoomeye-ai/mcp_zoomeye.git
cd mcp_zoomeye

# Build the Docker image
docker build -t zoomeyeteam/mcp-server-zoomeye:local .

# Run the container
docker run -i --rm -e ZOOMEYE_API_KEY=your_api_key_here zoomeyeteam/mcp-server-zoomeye:local

Using uv

uv is a fast Python package installer and resolver written in Rust. It's a modern alternative to pip that offers significant performance improvements.

Installation of uv

# Install uv using curl (macOS/Linux)
curl -LsSf https://astral.sh/uv/install.sh | sh

# Or using PowerShell (Windows)
irm https://astral.sh/uv/install.ps1 | iex

# Or using Homebrew (macOS)
brew install uv

Using uvx to run mcp-server-zoomeye

No specific installation is required when using uvx, which allows you to run Python packages directly:

Installing with uv

Alternatively, you can install the package using uv:

# Install in the current environment
uv pip install mcp-server-zoomeye

# Or create and install in a new virtual environment
uv venv
uv pip install mcp-server-zoomeye

Configuration

Environment Variables

The ZoomEye MCP server requires the following environment variable:

• ZOOMEYE_API_KEY: Your ZoomEye API key for authentication

You can set this environment variable in several ways:

1. Export in your shell session:

   export ZOOMEYE_API_KEY="your_api_key_here"
   

2. Pass directly when running the container (for Docker):

   docker run -i --rm -e ZOOMEYE_API_KEY=your_api_key_here zoomeyeteam/mcp-server-zoomeye:latest
   

Configure Claude.app

Add the following in Claude settings:

"mcpServers": {
  "zoomeye": {
    "command": "uvx",
    "args": ["mcp-server-zoomeye"],
    "env": {
        "ZOOMEYE_API_KEY": "your_api_key_here"
    }
  }
}

"mcpServers": {
  "zoomeye": {
    "command": "docker",
    "args": ["run", "-i", "--rm", "-e", "ZOOMEYE_API_KEY=your_api_key_here", "zoomeyeteam/mcp-server-zoomeye:latest"],
    "env": {
      "ZOOMEYE_API_KEY": "your_api_key_here"
    }
  }
}

"mcpServers": {
  "zoomeye": {
    "command": "python",
    "args": ["-m", "mcp_server_zoomeye"],
    "env": {
        "ZOOMEYE_API_KEY": "your_api_key_here"
    }
  }
}

Configure Zed

Add the following in Zed's settings.json:

"context_servers": [
  "mcp-server-zoomeye": {
    "command": "uvx",
    "args": ["mcp-server-zoomeye"],
    "env": {
        "ZOOMEYE_API_KEY": "your_api_key_here"
    }
  }
],

"context_servers": {
  "mcp-server-zoomeye": {
    "command": "python",
    "args": ["-m", "mcp_server_zoomeye"],
    "env": {
        "ZOOMEYE_API_KEY": "your_api_key_here"
    }
  }
},

Example Interactions

Example 1: Retrieve global Apache Tomcat assets

{
  "name": "zoomeye_search",
  "arguments": {
    "qbase64": "app=\"Apache Tomcat\""
  }
}

Response:

{
  "code": 60000,
  "message": "success",
  "total": 163139107,
  "query": "title=\"cisco vpn\"",
  "data": [
    {
      "url": "https://1.1.1.1:443",
      "ssl.jarm": "29d29d15d29d29d00029d29d29d29dea0f89a2e5fb09e4d8e099befed92cfa",
      "ssl.ja3s": "45094d08156d110d8ee97b204143db14",
      "iconhash_md5": "f3418a443e7d841097c714d69ec4bcb8",
      "robots_md5": "0b5ce08db7fb8fffe4e14d05588d49d9",
      "security_md5": "0b5ce08db7fb8fffe4e14d05588d49d9",
      "ip": "1.1.1.1",
      "domain": "www.google.com",
      "hostname": "SPACEX",
      "os": "windows",
      "port": 443,
      "service": "https",
      "title": ["GoogleGoogle appsGoogle Search"],
      "version": "1.1.0",
      "device": "webcam",
      "rdns": "c01031-001.cust.wallcloud.ch",
      "product": "OpenSSD",
      "header": "HTTP/1.1 302 Found Location: https://www.google.com/?gws_rd=ssl Cache-Control: private...",
      "header_hash": "27f9973fe57298c3b63919259877a84d",
      "body": "HTTP/1.1 302 Found Location: https://www.google.com/?gws_rd=ssl Cache-Control: private...",
      "body_hash": "84a18166fde3ee7e7c974b8d1e7e21b4",
      "banner": "SSH-2.0-OpenSSH_7.6p1 Ubuntu-4ubuntu0.3",
      "update_time": "2024-07-03T14:34:10",
      "header.server.name": "nginx",
      "header.server.version": "1.8.1",
      "continent.name": "Europe",
      "country.name": "Germany",
      "province.name": "Hesse",
      "city.name": "Frankfurt",
      "lon": "118.753262",
      "lat": "32.064838",
      "isp.name": "aviel.ru",
      "organization.name": "SERVISFIRST BANK",
      "zipcode": "210003",
      "idc": 0,
      "honeypot": 0,
      "asn": 4837,
      "protocol": "tcp",
      "ssl": "SSL Certificate Version: TLS 1.2 CipherSuit: TLS_ECDHE_RSA_WITH_AES_128_GCM_SHA256...",
      "primary_industry": "Finance",
      "sub_industry": "bank",
      "rank": 60
    }
  ]
}

Debugging and Troubleshooting

Using MCP Inspector

The Model Context Protocol Inspector is a tool that helps debug MCP servers by simulating client interactions. You can use it to test your ZoomEye MCP server:

# For uvx installation
npx @modelcontextprotocol/inspector uvx mcp-server-zoomeye

# If developing locally
cd path/to/servers/src/mcp_server_zoomeye
npx @modelcontextprotocol/inspector uv run mcp-server-zoomeye

Common Issues

1. Authentication Errors

   • Ensure your ZoomEye API key is correct and properly set as an environment variable
   • Check that your API key has not expired or been revoked

2. Connection Issues

   • Verify your internet connection
   • Check if the ZoomEye API is experiencing downtime

3. No Results

   • Your query might be too specific or contain syntax errors
   • Try simplifying your query or using different search terms

4. Rate Limiting

   • ZoomEye API has rate limits based on your account type
   • Space out your requests or upgrade your account for higher limits

Advanced Usage

Caching

The ZoomEye MCP server implements caching to improve performance and reduce API calls:

• Responses are cached based on the query parameters
• Cache duration is configurable (default: 1 hour)
• You can bypass the cache by setting ignore_cache to true in your query

Custom Fields

You can request specific fields in your query results by using the fields parameter:

{
  "name": "zoomeye_search",
  "arguments": {
    "qbase64": "app=\"Apache\"",
    "fields": "ip,port,domain,service,os,country,city"
  }
}

Pagination

For queries that return many results, you can paginate through them:

{
  "name": "zoomeye_search",
  "arguments": {
    "qbase64": "app=\"Apache\"",
    "page": 2,
    "pagesize": 20
  }
}

Contributing

We encourage contributions to mcp-server-zoomeye to help expand and improve its functionality. Whether it's adding new related tools, enhancing existing features, or improving documentation, your input is valuable.

For examples of other MCP servers and implementation patterns, see: https://github.com/modelcontextprotocol/servers

Pull requests are welcome! Feel free to contribute new ideas, bug fixes, or enhancements to make mcp-server-zoomeye more robust and practical.

License

mcp-server-zoomeye is licensed under the MIT License. This means you are free to use, modify, and distribute the software, subject to the terms and conditions of the MIT License. For more information, see the LICENSE file in the project repository.