FOFA Quake Hunter MCP Server

FOFA Quake Hunter MCP Server

An MCP server that enables AI models to query FOFA, 360 Quake, and Hunter cyberspace mapping platforms for asset discovery and security research. It supports natural language parameter configuration and provides comprehensive search tools for retrieving IP, port, and domain data.

Category
Visit Server

README

FOFA Quake Hunter MCP Server

一个用于查询 FOFA、Quake 和 Hunter 网络空间测绘平台的 MCP (Model Context Protocol) 服务器。

English | 中文


中文

功能特性

  • 🔍 FOFA 查询: 支持 FOFA 网络空间测绘系统的资产查询
  • 🌐 Quake 查询: 支持 360 Quake 网络空间测绘系统的深度查询
  • 🦅 Hunter 查询: 支持奇安信鹰图平台的资产查询
  • 🤖 AI 友好: 所有参数支持自然语言对话设置
  • ⚙️ 灵活配置: 可按需配置单个或多个平台

安装

使用 uvx (推荐)

uvx fofa-quake-hunter-mcp

使用 pip

pip install fofa-quake-hunter-mcp

从源码安装

git clone https://github.com/RuoJi6/fofa_quake_hunter_mcp.git
cd fofa_quake_hunter_mcp
pip install -e .

配置

在 MCP 客户端中配置

在 MCP 配置文件中添加(例如 Claude Desktop 的 claude_desktop_config.json 或 Kiro 的 .kiro/settings/mcp.json):

{
  "mcpServers": {
    "fofa-quake-hunter": {
      "command": "uvx",
      "args": ["fofa-quake-hunter-mcp"],
      "env": {
        "FOFA_EMAIL": "your_email@example.com",
        "FOFA_KEY": "your_fofa_api_key",
        "QUAKE_KEY": "your_quake_api_key",
        "HUNTER_KEY": "your_hunter_api_key"
      }
    }
  }
}

注意: 你可以只配置需要使用的平台,未配置的工具在调用时会返回友好的配置提示。

获取 API Key

功能说明

1. FOFA 查询 (fofa_search)

主要参数:

  • query: 查询语法(支持多种匹配和逻辑运算符)
  • size: 返回条数(默认 100,最大 10000)
  • page: 页码(默认 1)
  • fields: 返回字段(默认:host,ip,port,domain,title

匹配运算符:

  • = - 匹配(模糊匹配),=""时可查询不存在字段或值为空的情况
  • == - 完全匹配,==""时可查询存在且值为空的情况
  • != - 不匹配,!=""时可查询值为空的情况
  • *= - 模糊匹配,使用 *? 通配符(个人版及以上)

逻辑运算符:

  • && - 与(AND)
  • || - 或(OR)
  • () - 括号确认查询优先级

查询示例:

# 基础匹配
body="miner start"
domain="example.com"

# 完全匹配
server=="Microsoft-IIS/10"

# 不匹配
body="admin" && country!="CN"

# 模糊匹配(通配符)
banner*="mys??"

# 逻辑 AND(&&)
domain="example.com" && port="443"
title="login" && country="CN"

# 逻辑 OR(||)
title="admin" || title="后台"
port="80" || port="443"

# 优先级控制
(title="admin" || title="login") && country="CN"

2. Quake 查询 (quake_search)

主要参数:

  • query: 查询语法(使用冒号语法:field:value
  • size: 返回条数(默认 100)
  • include: 包含字段(逗号分隔,见下方可用字段列表)
  • exclude: 排除字段(逗号分隔)
  • pagination_id: 深度翻页 ID(5分钟有效)
  • start_time / end_time: 时间范围(UTC格式:2020-10-14 00:00:00)

查询语法:

  • 使用冒号 : 连接字段和值,如 port:443title:"keyword"
  • 逻辑运算符:ANDORNOT(大写)
  • 括号 () 控制优先级

可用字段(注册用户 - 服务数据):

ip, port, hostname, transport, asn, org, service.name, 
location.country_cn, location.province_cn, location.city_cn, 
service.http.host, service.http.title, service.http.server

可用字段(会员用户 - 额外服务数据字段):

time, domain, service.response, service.cert, 
components.product_catalog, components.product_type, 
components.product_level, components.product_vendor, 
location.country_en, location.province_en, location.city_en, 
location.district_en, location.district_cn, location.isp, 
service.http.body, components.product_name_cn, components.version, 
service.http.infomation.mail, service.http.favicon.hash, 
service.http.favicon.data, service.http.status_code

查询示例:

# 单条件查询
port:443
title:"后台管理"

# 逻辑 AND
port:3389 AND country:"China"
port:3389 AND country_cn:"中国" AND NOT province_cn:"广东"

# 逻辑 OR
port:3389 AND (country:"China" OR country:"United States")

# 逻辑 NOT
port:80 AND NOT response:"baidu"
service:http AND NOT response:"baidu"

# 优先级控制
port:3389 AND (country:"China" OR country:"United States") AND NOT province_cn:"广东"

字段筛选示例:

# 只返回 IP 和端口
include: "ip,port"

# 返回 IP、端口和网页标题
include: "ip,port,service.http.title"

# 返回基础信息和组织
include: "ip,port,service.http.title,org,asn"

# 返回完整信息(会员)- 注意使用具体的组件字段
include: "ip,port,service.http.title,service.http.server,domain,components.product_name_cn,components.version"

⚠️ 常见字段错误:

  • components → ✅ 使用具体字段如 components.product_name_cn
  • as_org → ✅ 使用 asnorg
  • as_organization → ✅ 使用 asnorg

3. Hunter 查询 (hunter_search)

主要参数:

  • query: 查询语法(使用等号语法:field="value"
  • page_size: 每页条数(可选:10/50/100,默认 10)
  • page: 页码(默认 1)
  • is_web: 资产类型(1=web资产,2=非web资产,3=全部)
  • fields: 返回字段
  • start_time / end_time: 时间范围(格式:YYYY-MM-DD)

匹配运算符:

  • = - 模糊查询,查询包含关键词的资产
  • == - 精确查询,查询有且仅有关键词的资产
  • != - 模糊剔除,剔除包含关键词的资产。使用 !="" 可查询值不为空的情况
  • !== - 精确剔除,剔除有且仅有关键词的资产

逻辑运算符:

  • && - 与(AND)
  • || - 或(OR)
  • () - 括号内表示查询优先级最高

查询示例:

# 模糊匹配
web.body="keyword"
domain="example.com"

# 精确匹配
web.title=="登录"
ip=="1.1.1.1"

# 模糊剔除
web.body="admin" && ip!="1.1.1.1"

# 精确剔除
domain!=="example.com"

# 查询值不为空
ip!=""

# 逻辑 AND(&&)
web.title="后台管理系统" && ip="1.1.1.1"
domain="example.com" && web.status_code="200"

# 逻辑 OR(||)
domain="example.com" || domain="test.com"
web.title="admin" || web.title="login"

# 优先级控制
(web.title="admin" || web.title="login") && ip!=""

AI 对话示例

用户: 查询 FOFA,body="admin",返回 50 条
AI: 将设置 query="body=\"admin\"", size=50

用户: 查询 Quake,标题为"后台管理",只返回 IP 和端口
AI: 将设置 query='title:"后台管理"', include='ip,port'

用户: 查询 Hunter,web.title="登录",只要 web 资产,每页 100 条
AI: 将设置 query='web.title="登录"', is_web=1, page_size=100

功能对比

功能 FOFA Quake Hunter
返回条数控制 ✅ size (1-10000) ✅ size (1-500) ✅ page_size (10/50/100)
字段控制 ✅ fields ✅ include/exclude ✅ fields
翻页 ✅ page ✅ pagination_id ✅ page
时间范围 ✅ start_time/end_time ✅ start_time/end_time
资产类型筛选 ✅ is_web

开发

# 克隆仓库
git clone https://github.com/RuoJi6/fofa_quake_hunter_mcp.git
cd fofa_quake_hunter_mcp

# 安装依赖
uv sync

# 运行服务器
uv run fofa-quake-hunter-mcp

许可证

MIT License - 详见 LICENSE 文件

贡献

欢迎提交 Issue 和 Pull Request!


English

Features

  • 🔍 FOFA Search: Query FOFA cyberspace mapping platform
  • 🌐 Quake Search: Query 360 Quake cyberspace mapping platform with deep pagination
  • 🦅 Hunter Search: Query Qianxin Hunter (鹰图) platform
  • 🤖 AI-Friendly: All parameters support natural language configuration
  • ⚙️ Flexible Config: Configure only the platforms you need

Installation

Using uvx (Recommended)

uvx fofa-quake-hunter-mcp

Using pip

pip install fofa-quake-hunter-mcp

From Source

git clone https://github.com/RuoJi6/fofa_quake_hunter_mcp.git
cd fofa_quake_hunter_mcp
pip install -e .

Configuration

Configure in MCP Client

Add to your MCP configuration file (e.g., Claude Desktop's claude_desktop_config.json or Kiro's .kiro/settings/mcp.json):

{
  "mcpServers": {
    "fofa-quake-hunter": {
      "command": "uvx",
      "args": ["fofa-quake-hunter-mcp"],
      "env": {
        "FOFA_EMAIL": "your_email@example.com",
        "FOFA_KEY": "your_fofa_api_key",
        "QUAKE_KEY": "your_quake_api_key",
        "HUNTER_KEY": "your_hunter_api_key"
      }
    }
  }
}

Note: You can configure only the platforms you need. Unconfigured tools will show friendly setup instructions when called.

Get API Keys

Tools

1. FOFA Search (fofa_search)

Key Parameters:

  • query: Search query (e.g., body="admin", domain="example.com")
  • size: Number of results (default: 100, max: 10000)
  • page: Page number (default: 1)
  • fields: Fields to return (default: host,ip,port,domain,title)

Query Examples:

body="miner start"
domain="example.com" && port="443"
title="login" && country="CN"

2. Quake Search (quake_search)

Key Parameters:

  • query: Search query (e.g., title:"admin", ip:1.1.1.1)
  • size: Number of results (default: 100)
  • include: Fields to include (e.g., ip,port,service.http.title)
  • exclude: Fields to exclude
  • pagination_id: Pagination ID for deep paging (5-minute expiry)
  • start_time / end_time: Time range (UTC format)

Query Examples:

title:"admin panel"
ip:1.1.1.1 AND port:80
service:http AND country:"china"

3. Hunter Search (hunter_search)

Key Parameters:

  • query: Search query (e.g., web.body="admin", ip="1.1.1.1")
  • page_size: Results per page (options: 10/50/100, default: 10)
  • page: Page number (default: 1)
  • is_web: Asset type (1=web assets, 2=non-web assets, 3=all)
  • fields: Fields to return
  • start_time / end_time: Time range (format: YYYY-MM-DD)

Query Examples:

web.body="keyword"
web.title="admin panel"
domain="example.com" && web.status_code="200"

Feature Comparison

Feature FOFA Quake Hunter
Result Count ✅ size (1-10000) ✅ size (1-500) ✅ page_size (10/50/100)
Field Control ✅ fields ✅ include/exclude ✅ fields
Pagination ✅ page ✅ pagination_id ✅ page
Time Range ✅ start_time/end_time ✅ start_time/end_time
Asset Type Filter ✅ is_web

Development

# Clone repository
git clone https://github.com/RuoJi6/fofa_quake_hunter_mcp.git
cd fofa_quake_hunter_mcp

# Install dependencies
uv sync

# Run server
uv run fofa-quake-hunter-mcp

License

MIT License - see LICENSE file for details

Contributing

Issues and Pull Requests are welcome!

Links

Recommended Servers

playwright-mcp

playwright-mcp

A Model Context Protocol server that enables LLMs to interact with web pages through structured accessibility snapshots without requiring vision models or screenshots.

Official
Featured
TypeScript
Magic Component Platform (MCP)

Magic Component Platform (MCP)

An AI-powered tool that generates modern UI components from natural language descriptions, integrating with popular IDEs to streamline UI development workflow.

Official
Featured
Local
TypeScript
Audiense Insights MCP Server

Audiense Insights MCP Server

Enables interaction with Audiense Insights accounts via the Model Context Protocol, facilitating the extraction and analysis of marketing insights and audience data including demographics, behavior, and influencer engagement.

Official
Featured
Local
TypeScript
VeyraX MCP

VeyraX MCP

Single MCP tool to connect all your favorite tools: Gmail, Calendar and 40 more.

Official
Featured
Local
graphlit-mcp-server

graphlit-mcp-server

The Model Context Protocol (MCP) Server enables integration between MCP clients and the Graphlit service. Ingest anything from Slack to Gmail to podcast feeds, in addition to web crawling, into a Graphlit project - and then retrieve relevant contents from the MCP client.

Official
Featured
TypeScript
Kagi MCP Server

Kagi MCP Server

An MCP server that integrates Kagi search capabilities with Claude AI, enabling Claude to perform real-time web searches when answering questions that require up-to-date information.

Official
Featured
Python
E2B

E2B

Using MCP to run code via e2b.

Official
Featured
Neon Database

Neon Database

MCP server for interacting with Neon Management API and databases

Official
Featured
Exa Search

Exa Search

A Model Context Protocol (MCP) server lets AI assistants like Claude use the Exa AI Search API for web searches. This setup allows AI models to get real-time web information in a safe and controlled way.

Official
Featured
Qdrant Server

Qdrant Server

This repository is an example of how to create a MCP server for Qdrant, a vector search engine.

Official
Featured