Xuanmu-BugBounty-mcp
独立自包含的SRC挖洞专用MCP工具包,提供32个侦察、漏洞检测、认证安全、信息提取等工具,零外部依赖,即装即用。
README
🎯 Xuanmu-BugBounty-mcp
玄幕安全团队 · guaidao2 开发
独立自包含的 SRC 挖洞专用 MCP 工具包 — 零外部依赖,即装即用
📦 快速安装
# 1. 克隆仓库
git clone https://github.com/guaidao2/Xuanmu-Bugbounty-mcp.git
cd Xuanmu-BugBounty-mcp
# 2. 安装依赖(使用清华源)
pip install -e . -i https://pypi.tuna.tsinghua.edu.cn/simple
# 3. 启动 MCP 服务器(stdio 模式)
# 方式一:直接启动(推荐)
xuanmu-bb
# 方式二:使用 python + 脚本绝对路径启动
python "绝对路径/src/xuanmu_bb/server.py"
🔌 MCP 配置
在支持 MCP 的客户端(如 Claude Desktop、Cursor、Windsurf 等)中配置:
方式一:使用 xuanmu-bb 命令(推荐)
{
"mcpServers": {
"Xuanmu-BugBounty-mcp": {
"command": "xuanmu-bb",
"args": []
}
}
}
方式二:使用 python + 脚本文件绝对路径调用
{
"mcpServers": {
"Xuanmu-BugBounty-mcp": {
"command": "python",
"args": ["你的项目绝对路径/src/xuanmu_bb/server.py"]
}
}
}
💡 将
你的项目绝对路径替换为你实际存放项目的完整路径 Windows 示例:"D:\\Projects\\Xuanmu-BugBounty-mcp\\src\\xuanmu_bb\\server.py"
🔐 认证方式(两种都支持)
目标需要登录?所有 HTTP 工具都支持两种认证:
| 方式 | 参数 | 适用场景 |
|---|---|---|
| Bearer Token | auth_token="xxx" |
JWT / OAuth2 令牌登录的系统 |
| Session Cookie | cookie="SESSION=xxx" |
表单登录 / PHP/Java Session 的系统 |
# Bearer Token(API 认证)
bb_sqli url="https://target.com/api/search" params="q" auth_token="eyJhbGciOiJIUzI1NiIs..."
# Session Cookie(Web 认证)
bb_xss url="https://target.com/search" params="q" cookie="JSESSIONID=ABCD1234"
# 先探测认证状态
bb_param_discover url="https://target.com"
# 输出会标注: [AUTH: required] HTTP 401 或 [AUTH: none] HTTP 200
🧰 全部 39 个工具
🔍 侦察模块 (Reconnaissance)
| 工具 | 功能 | 参数 |
|---|---|---|
bb_ping |
存活探测 — TCP + HTTP 双重检测 | target, timeout, proxy |
bb_port_scan |
端口扫描 — Top100/自定义范围 | target, ports, timeout, concurrent |
bb_subdomain |
子域名枚举 — DNS 批量解析 | domain, wordlist, concurrent |
bb_fingerprint |
Web 指纹识别 — 技术栈/CMS/WAF | url, proxy, cookie, auth_token |
bb_dir_scan |
目录爆破 — 内置 150+ 敏感路径 | url, wordlist, status_filter, concurrent, cookie, auth_token |
🔥 漏洞检测模块 (Vulnerability Detection)
| 工具 | 功能 | 参数 |
|---|---|---|
bb_sqli |
SQL 注入 — 报错/布尔/时间盲注 | url, params, method, body, auth_token, cookie, waf_mode |
bb_nosqli |
NoSQL 注入 — MongoDB $ne/$gt/$regex | url, params, method, body, auth_token, cookie |
bb_xss |
XSS 检测 — 反射型/多种上下文 | url, params, method, body, auth_token, cookie, waf_mode |
bb_ssti |
SSTI 模板注入 — 多引擎检测 | url, params, method, body, auth_token, cookie, waf_mode |
bb_cmdi |
命令注入 — 时间盲注+输出回显 | url, params, method, body, auth_token, cookie, waf_mode |
bb_ssrf |
SSRF 检测 — 内网/云元数据/OOB | url, params, method, body, auth_token, cookie, waf_mode |
bb_cors |
CORS 跨域 — 12 种 Origin 测试 | url, auth_token, cookie |
bb_open_redirect |
开放重定向 — 参数扫描+跳转测试 | url, params, auth_token, cookie |
bb_file_upload |
文件上传绕过 — 实际上传+验证可访问 | url, auth_token, cookie |
bb_csrf |
CSRF 检测 — Token/SameSite/Referer | url, auth_token, cookie |
bb_xxe |
XXE 检测 — 经典/Blind/SVG/XInclude | url, body, auth_token, cookie |
bb_lfi |
路径遍历 — ../遍历/PHP filter | url, params, method, body, auth_token, cookie, waf_mode |
bb_host_inject |
Host 头注入 — 9 种攻击场景 | url, auth_token, cookie |
bb_takeover |
子域名接管 — CNAME+50+云服务匹配 | domain, auth_token |
bb_race |
条件竞争 — 并发请求+响应差异分析 | url, method, data/body, concurrent, auth_token, cookie |
🔐 认证安全模块 (Authentication Security)
| 工具 | 功能 | 参数 |
|---|---|---|
bb_jwt_decode |
JWT 解码 — 解析 Header/Payload | token |
bb_jwt_analyze |
JWT 安全分析 — 漏洞检测+攻击建议 | token |
bb_jwt_crack |
JWT 密钥爆破 — HMAC 字典攻击 | token, wordlist |
bb_jwt_attack |
JWT 攻击 — None/KID注入/算法混淆 | token, mode, payload_override, public_key |
bb_graphql |
GraphQL 扫描 — Introspection/批量/递归 | url, auth_token, cookie |
📋 信息提取模块 (Information Extraction)
| 工具 | 功能 | 参数 |
|---|---|---|
bb_extract |
URL/Endpoint 提取 — HTML+JS分析 | url, depth, auth_token, cookie |
bb_secrets |
敏感信息检测 — 20 种正则模式 | url, auth_token, cookie, check_js |
bb_headers |
安全头审计 — 8 项评分+修复建议 | url, auth_token, cookie |
bb_param_discover |
参数自动发现 — 提取表单/查询/JSON/JS | url, depth, auth_token, cookie |
bb_js_analyze |
JS 深度分析 — API路由/Sourcemap/硬编码/SPA/WebSocket | url, auth_token, cookie |
🧰 工具模块 (Utilities)
| 工具 | 功能 | 参数 |
|---|---|---|
bb_send |
手工 HTTP 发包 — 自定义方法/头/Body | url, method, headers, body, auth_token, cookie |
bb_payload |
Payload 工厂 — 9 类漏洞×6 种编码 | vuln_type, encode, count |
bb_oob |
OOB 外带辅助 — 生成回调标识/Payload | mode, callback_url |
bb_idor |
IDOR 越权检测 — 双 Token 对比 + 序号枚举 | url, token_owner, token_attacker, method, param |
bb_cloud_check |
云服务安全 — S3/元数据/云配置泄露 | url, auth_token, cookie |
bb_waf_check |
WAF 指纹识别 — 14 种 WAF + 绕过建议 | url, auth_token, cookie |
bb_session |
多步骤流程 — Session保持 + 请求链 | steps (JSON), proxy |
bb_report |
漏洞报告生成 — SRC 格式 Markdown | vuln_type, target, param, payload, impact |
🚀 使用示例
# 侦察一个目标
bb_ping target="example.com"
bb_port_scan target="example.com" ports="80,443,8080-8090"
bb_subdomain domain="example.com"
bb_fingerprint url="https://example.com"
bb_dir_scan url="https://example.com"
# 漏洞检测
bb_sqli url="https://example.com/page?id=1" params="id"
bb_xss url="https://example.com/search?q=test" params="q"
bb_ssti url="https://example.com/welcome?name=test" params="name"
bb_cmdi url="https://example.com/ping?host=127.0.0.1" params="host"
bb_ssrf url="https://example.com/fetch?url=http://example.com" params="url"
bb_cors url="https://api.example.com/data"
bb_open_redirect url="https://example.com/redirect?url=https://example.com"
bb_csrf url="https://example.com/change_password"
bb_xxe url="https://example.com/xml/parse"
bb_lfi url="https://example.com/file?name=test.txt" params="name"
bb_host_inject url="https://example.com/reset"
bb_takeover domain="sub.example.com"
bb_race url="https://example.com/api/coupon/claim" method="POST"
# JWT 工具
bb_jwt_decode token="eyJhbGciOiJIUzI1NiIs..."
bb_jwt_analyze token="eyJhbGciOiJIUzI1NiIs..."
bb_jwt_crack token="eyJhbGciOiJIUzI1NiIs..."
bb_jwt_attack token="eyJhbGciOiJIUzI1NiIs..." mode="none"
# GraphQL
bb_graphql url="https://example.com/graphql"
# 信息提取
bb_extract url="https://example.com" depth=2
bb_secrets url="https://example.com" check_js=true
bb_headers url="https://example.com"
# POST body 注入(新增)
bb_sqli url="https://target.com/api/login" method="POST" body="username=admin&password=test" params="username"
bb_xss url="https://target.com/api/feedback" method="POST" body="message=test" params="message"
bb_cmdi url="https://target.com/api/ping" method="POST" body="host=127.0.0.1" params="host"
bb_xxe url="https://target.com/xml/parse" body="<?xml version='1.0'?><root>test</root>"
bb_lfi url="https://target.com/api/read" method="POST" body="file=../../../../etc/passwd" params="file"
bb_ssti url="https://target.com/welcome" method="POST" body="name={{7*7}}" params="name"
# 工具
bb_payload vuln_type="xss" encode="all" count=20
bb_send url="https://example.com/api/login" method="POST" headers='{"Content-Type": "application/json"}' body='{"user":"admin","pass":"123"}'
bb_report vuln_type="sqli" target="https://example.com/page?id=1" payload="\' OR 1=1 -- "
bb_send url="https://example.com/api/login" method="POST" headers='{"Content-Type": "application/json"}' body='{"user":"admin","pass":"123"}'
bb_report vuln_type="sqli" target="https://example.com/page?id=1" payload="' OR 1=1 -- "
# JS 深度分析
bb_js_analyze url="https://example.com" auth_token="..."
# WAF 检测
bb_waf_check url="https://example.com"
# IDOR 越权检测
bb_idor url="https://api.example.com/users/1234" \
token_owner="eyJ...owner_token..." \
token_attacker="eyJ...attacker_token..."
# 多步骤流程
bb_session steps='[
{"method":"POST","url":"https://target.com/login","body":"user=admin&pass=test"},
{"method":"GET","url":"https://target.com/api/profile"}
]'
# 云服务检测
bb_cloud_check url="https://example.com"
# 参数自动发现 + NoSQL 注入
bb_param_discover url="https://example.com/api"
bb_nosqli url="https://example.com/api/user?id=1"
🛡️ WAF 防护引擎
支持 WAF 检测的扫描工具(sqli/xss/ssti/cmdi/ssrf/lfi/dir_scan)自动集成:
| 参数 | 默认值 | 说明 |
|---|---|---|
waf_mode |
safe |
off(不管)/ safe(降速+轮换UA)/ aggressive(尝试绕过) |
max_retries_on_block |
3 |
被拦截后重试次数,超限自动熔断 |
request_delay |
auto |
auto=检测到WAF后从0.5s升到3s,也可手动指定秒数 |
工作流程:
1. 预检 — 发无害请求检测 WAF 指纹
2. 发现 WAF → 自动降速 + 轮换 UA + 提示绕过建议
3. 扫描中 — 实时检测拦截页面 (403/503/验证页)
4. 熔断 — 连续 N 次被拦截 → 自动中断 → 输出熔断报告
先检测 WAF:
bb_waf_check url="https://target.com"
带 WAF 防护扫描:
bb_sqli url="https://target.com/page?id=1" waf_mode="safe" request_delay="3"
🏗️ 项目结构
src/xuanmu_bb/
├── server.py # MCP 入口(38 个工具注册)
├── client.py # HTTP 客户端(代理/Cookie/UA轮换/反封策略)
├── utils.py # 公共工具函数
├── data/ # 内置数据(Payload 字典/指纹/WAF库/正则模式)
├── recon/ # 侦察模块(5 工具)
├── vuln/ # 漏洞检测模块(15 工具)
├── auth/ # 认证安全模块(5 工具)
├── extract/ # 信息提取模块(5 工具)
└── tools/ # 工具模块(9 工具)
🧪 技术特点
- 完全自包含 — 不依赖 nmap/nuclei/burp 等外部工具,纯 Python 实现
- 零外部扫描器依赖 — 不依赖 yakit/tscanplus/nuclei,独立运行
- 内置丰富数据 — 200+ 子域字典 / 150+ 目录字典 / 1000+ 端口服务映射 / 500+ Payload
- 反封策略 — User-Agent 轮换 / 请求间隔控制 / 代理支持
- 清华源 — 默认使用
pypi.tuna.tsinghua.edu.cn加速安装 - 覆盖 SRC 全流程 — 信息收集 → 漏洞检测 → 利用验证 → 报告生成
📝 许可证
本项目仅供合法的安全测试与漏洞挖掘使用。使用者需遵守相关法律法规。
Xuanmu-BugBounty-mcp © 2026 玄幕安全团队 · guaidao2
Recommended Servers
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.
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.
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.
VeyraX MCP
Single MCP tool to connect all your favorite tools: Gmail, Calendar and 40 more.
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.
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.
E2B
Using MCP to run code via e2b.
Neon Database
MCP server for interacting with Neon Management API and databases
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.
Qdrant Server
This repository is an example of how to create a MCP server for Qdrant, a vector search engine.