Android Proxy MCP
Allows capturing and analyzing HTTP/HTTPS network traffic from Android devices, providing tools to search request/response content, filter traffic by domain or status code, and manage large response payloads for detailed analysis.
Click on "Deploy Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type
@followed by the MCP server name and your instructions, e.g., "@Android Proxy MCP Show me the most recent failed requests and analyze the error responses."
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
Android Proxy MCP
基于 MCP (Model Context Protocol) 的 Android 抓包服务,让 AI 助手能够帮你抓取和分析 HTTP/HTTPS 流量。
功能特点
抓包: 捕获 HTTP/HTTPS 流量,支持按域名、状态码、资源类型筛选
智能搜索: 搜索请求/响应内容,支持大响应分片读取
AI 驱动: 通过自然语言让 Claude 帮你分析网络请求
Related MCP server: httptoolkit-mcp
架构
┌─────────────────┐ SQLite ┌─────────────────┐
│ 代理服务 │ ─────────────→ │ MCP 服务 │
│ (终端手动启动) │ 流量数据共享 │ (Claude 调用) │
│ mitmdump │ │ 查询/搜索/分析 │
└─────────────────┘ └─────────────────┘
↑
│ HTTP/HTTPS
│
┌─────────┐
│ 手机 │
└─────────┘快速开始
1. 环境要求
Python 3.11+
uv (Python 包管理器)
安装 uv:
# macOS / Linux
curl -LsSf https://astral.sh/uv/install.sh | sh
# Windows (PowerShell)
powershell -ExecutionPolicy ByPass -c "irm https://astral.sh/uv/install.ps1 | iex"
# 或通过 pip
pip install uv2. 安装
# 克隆项目
git clone https://github.com/yourname/android-proxy-mcp.git
cd android-proxy-mcp
# 安装依赖
uv sync3. 配置 Claude Desktop
编辑 Claude Desktop 配置文件:
macOS: ~/Library/Application Support/Claude/claude_desktop_config.json
Windows: %APPDATA%\Claude\claude_desktop_config.json
{
"mcpServers": {
"android-proxy": {
"command": "uv",
"args": ["--directory", "/path/to/android-proxy-mcp", "run", "android-proxy-mcp"]
}
}
}将
/path/to/android-proxy-mcp替换为实际项目路径
4. 重启 Claude Desktop
配置完成后,重启 Claude Desktop 使配置生效。
使用方法
第一步:启动代理
在终端中运行:
uv run android-proxy-start你会看到如下输出:
╔════════════════════════════════════════════════════════════╗
║ 🚀 Android Proxy MCP 启动向导 ║
╚════════════════════════════════════════════════════════════╝
════════════════════════════════════════════════════════════
环境检测
════════════════════════════════════════════════════════════
✓ 端口 8288 可用
════════════════════════════════════════════════════════════
手机配置
════════════════════════════════════════════════════════════
手机 Wi-Fi 代理设置:
┌─────────────────────────────────┐
│ 服务器: 192.168.1.100 │
│ 端 口: 8288 │
└─────────────────────────────────┘
证书安装: 手机浏览器访问 http://mitm.it
════════════════════════════════════════════════════════════
启动代理 (Ctrl+C 停止)
════════════════════════════════════════════════════════════保持这个终端窗口运行,不要关闭。
第二步:配置手机代理
确保手机和电脑在同一 Wi-Fi 网络下。
手机进入 设置 → Wi-Fi
长按当前连接的 Wi-Fi → 修改网络
展开 高级选项
代理设置选择 手动
填写启动脚本显示的 IP 和端口
保存
第三步:安装 CA 证书(抓 HTTPS 必须)
手机浏览器访问
http://mitm.it(通过代理访问)选择 Android 图标下载证书
设置 → 安全 → 加密与凭据 → 安装证书 → CA 证书
选择下载的证书文件安装
⚠️ 注意: Android 7+ 用户证书默认不被 App 信任,只能抓取浏览器和部分应用的 HTTPS 流量。
安装系统证书(Root 用户)
如果你的设备已 Root(安装了 Magisk),可以将用户证书移动为系统证书:
先按照上述步骤安装用户证书
安装
resources/MoveCertificate-v1.5.5.zipMagisk 模块:打开 Magisk → 模块 → 从本地安装
选择项目中的
resources/MoveCertificate-v1.5.5.zip重启设备
重启后,用户证书会自动移动到系统证书目录
第四步:在 Claude 中查询流量
打开 Claude Desktop,用自然语言查询流量:
基础查询:
"显示最近的网络请求" "显示 api.example.com 的请求" "显示所有失败的请求(状态码 4xx 或 5xx)"
搜索内容:
"搜索响应中包含 '张三' 的请求" "搜索 URL 中包含 search 的请求" "搜索请求头中包含 X-Token 的请求"
查看大响应:
"读取 req-5 的响应体" "继续读取 req-5 响应体,从 4000 开始"
智能分析:
"帮我找酷安 app 的搜索接口" "分析这个 API 的请求参数"
第五步:停止抓包
在运行代理的终端窗口按 Ctrl+C 停止代理。
记得在手机 Wi-Fi 设置中关闭代理。
MCP 工具列表
工具 | 说明 |
| 获取代理状态 |
| 列出流量(支持域名/状态码/类型筛选) |
| 搜索流量内容(URL/请求头/请求体/响应头/响应体) |
| 获取请求元数据(请求头、响应头等) |
| 分片读取大响应体 |
| 清空流量记录 |
| 获取证书安装指南 |
常见问题
Q: 手机配置代理后无法上网?
确认电脑和手机在同一 Wi-Fi
确认代理已启动(检查终端窗口)
检查电脑防火墙是否允许 8288 端口
尝试用电脑 IP(不是 localhost)
Q: 能抓到 HTTP 但抓不到 HTTPS?
需要安装 CA 证书。手机浏览器访问 http://mitm.it 下载安装。
Q: 安装了证书但某些 App 还是抓不到 HTTPS?
Android 7+ 用户证书默认不被 App 信任 → 需要安装系统证书(参考上方)
部分 App 有 SSL Pinning(证书锁定)→ 参考下方绕过方案
Q: 如何绕过 SSL Pinning?
推荐使用 LSPosed + 禁用 SSL 验证插件:
Q: 响应太大,MCP 无法返回?
使用 traffic_search 搜索关键词定位,然后用 traffic_read_body 分片读取。
项目结构
android-proxy-mcp/
├── README.md
├── pyproject.toml
├── src/
│ └── android_proxy_mcp/
│ ├── cli/ # 命令行工具
│ │ └── start.py # 代理启动脚本
│ ├── core/ # 核心模块
│ │ └── sqlite_store.py # SQLite 流量存储
│ ├── tools/ # MCP 工具
│ └── server.py # MCP 服务入口
├── tests/
├── docs/ # 文档
└── resources/ # 资源文件
└── MoveCertificate-v1.5.5.zip # 证书移动模块开发
# 安装开发依赖
uv sync --extra dev
# 运行测试
uv run pytest tests/ -v
# 代码格式化
uv run ruff format .许可证
MIT License
Available Tools
11 toolsandroid_clear_proxyC
清除 Android 设备上的代理设置
Inputs: serial (required).
android_get_device_infoB
获取指定 Android 设备的详细信息(型号、版本、是否 root 等)
Inputs: serial (required).
android_list_devicesB
列出所有连接的 Android 设备
Inputs: no named parameters.
android_setup_proxyA
在 Android 设备上设置 HTTP 代理。注意:此方式对部分应用可能无效,建议在 Wi-Fi 设置中手动配置。
Inputs: serial (required), proxy_host (required), proxy_port.
get_cert_infoA
获取 CA 证书信息和安装指南。抓取 HTTPS 流量需要在设备上安装此证书。
Inputs: no named parameters.
proxy_statusA
获取代理服务器状态。注意:需要先在终端运行 'uv run android-proxy-start' 启动代理。
Inputs: no named parameters.
traffic_clearB
清空所有捕获的流量
Inputs: no named parameters.
traffic_get_detailA
获取单个请求的元数据(请求头、响应头、参数等)。注意:不包含请求体和响应体内容,使用 traffic_read_body 读取。
Inputs: request_id (required).
traffic_listB
列出捕获的 HTTP/HTTPS 流量。支持按域名、类型、状态码、URL 筛选。
Inputs: limit, filter_domain, filter_type, filter_status, +1 more.
traffic_read_bodyB
分片读取请求体或响应体。用于查看大内容,支持分页读取。
Inputs: request_id (required), field, offset, length.
traffic_searchC
搜索流量内容。可搜索 URL、请求头、请求体、响应头、响应体。返回匹配的片段而非完整内容。
Inputs: keyword (required), search_in, method, domain, +2 more.
Tool Schema Changelog
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
Loading tool schema changelog…
TDQS
Scored across 11 tools
Each tool has a distinct purpose with clear boundaries: device management (list, get, clear/set proxy), certificate handling (get_cert_info), proxy server control (proxy_status), and traffic analysis (clear, list, get_detail, read_body, search). The descriptions explicitly differentiate overlapping tools like traffic_get_detail (metadata only) vs traffic_read_body (body content), preventing misselection.
Tool names follow a consistent snake_case pattern with clear prefixes: 'android_' for device operations, 'get_'/'traffic_' for certificate and traffic functions, and descriptive verb_noun combinations (e.g., traffic_list, traffic_search). No mixing of conventions or vague names, making the set predictable and readable.
11 tools are well-scoped for the server's purpose of Android proxy management and traffic analysis. The count covers device handling, proxy setup, certificate guidance, and comprehensive traffic inspection (listing, filtering, searching, reading details/bodies), with each tool earning its place without bloat or gaps.
The toolset provides complete coverage for the Android proxy domain: device discovery and info, proxy configuration (clear/set), certificate setup guidance, proxy server status, and full traffic lifecycle (capture via proxy, list, filter, search, view metadata, read bodies, clear). No obvious gaps exist; agents can perform end-to-end workflows without dead ends.
Maintenance
Related MCP Connectors
Capture, inspect & debug HTTPS traffic across iOS, Android, browsers & backends — 304 MCP tools.
MCP server that lets your AI assistant build and deploy apps, hosting each one at a live URL.
MCP server for building and testing AI agents with multi-model experimentation and insights.
MCP server for secureFlows: token-free URL builders and integration-linting tools for AI agents.
Related MCP Servers
- AlicenseAqualityFmaintenanceAn MCP server that enables AI assistants to control HTTP Toolkit for intercepting, inspecting, and debugging HTTP(S) traffic from browsers, mobile devices, and Docker containers. It provides tools for server management, interceptor activation, and sending HTTP requests through natural language commands.2318 npm1MIT
- AlicenseAqualityDmaintenanceMCP server for intercepting and mocking HTTP(S) traffic via a Mockttp proxy, with tools for Android emulator setup, traffic inspection, protobuf analysis, and rule-based manipulation.3518 npmMIT
- AlicenseAqualityDmaintenanceAn MCP server for parsing and analyzing HAR files, enabling AI assistants to inspect network traffic with automatic redaction of sensitive authentication headers.7MIT
- AlicenseNot gradedqualityDmaintenanceAn MCP server that leverages Wireshark/Tshark for real-time and offline network traffic analysis, enabling AI assistants to capture, parse, and detect threats with automated scanning and threat intelligence.2MIT