Fridare

Interactive Android Reverse Engineering through Frida x MCP

Attach once, explore forever. No scripts to restart, no context lost.

English | 中文

Features

Persistent sessions — attach once, run unlimited commands
22 MCP tools — full Frida lifecycle from any MCP-compatible client
r0capture integration — universal SSL/HTTP traffic capture with PCAP export
SSL unpinning — OkHttp / Conscrypt / WebView / universal bypass
Certificate dump — auto-export client certs as P12
Cross-platform — Windows, Linux, macOS. Uses your system Frida — no version conflicts

Install

Requires frida and frida-tools to be available in your active Python environment.

From GitHub

pip install git+https://github.com/MyuriKanao/fridare.git

Or install it as a standalone tool:

uv tool install git+https://github.com/MyuriKanao/fridare.git

For local development

pip install -e .

Publishing

For maintainers, build and publish the package with standard Python packaging tools:

python -m pip install --upgrade build twine
python -m build
python -m twine check dist/*
python -m twine upload dist/*

If you want to test the release flow first, upload to TestPyPI before publishing to PyPI.

GitHub Actions release flow

This repository also includes .github/workflows/pypi-publish.yml:

pull requests / pushes to main build the package and run twine check
pushes of tags matching v* publish to PyPI automatically

Typical release flow:

git tag v0.1.0
git push origin v0.1.0

Before the first automated release, configure a PyPI Trusted Publisher for:

PyPI project: fridare
GitHub owner: MyuriKanao
Repository: fridare
Workflow file: .github/workflows/pypi-publish.yml
Environment: pypi

If the PyPI project does not exist yet, create a pending publisher first and let the first tagged release create it.

Configure

Fridare is a standard stdio MCP server. Add it to any MCP-compatible client:

Claude Code (CLI / Desktop / Web)

claude mcp add fridare -- fridare

Or manually in ~/.claude.json:

{
  "mcpServers": {
    "fridare": {
      "type": "stdio",
      "command": "fridare"
    }
  }
}

Claude Desktop

Add to claude_desktop_config.json (Settings > Developer > Edit Config):

{
  "mcpServers": {
    "fridare": {
      "command": "fridare"
    }
  }
}

If installed in a virtualenv, use the full path:

{
  "mcpServers": {
    "fridare": {
      "command": "/path/to/venv/bin/fridare"
    }
  }
}

Cursor

Add to .cursor/mcp.json in your project root (or global settings):

{
  "mcpServers": {
    "fridare": {
      "command": "fridare"
    }
  }
}

VS Code (GitHub Copilot)

Add to .vscode/mcp.json:

{
  "servers": {
    "fridare": {
      "type": "stdio",
      "command": "fridare"
    }
  }
}

Windsurf

Add to ~/.codeium/windsurf/mcp_config.json:

{
  "mcpServers": {
    "fridare": {
      "command": "fridare"
    }
  }
}

Codex CLI (OpenAI)

Add to ~/.codex/config.json:

{
  "mcpServers": {
    "fridare": {
      "type": "stdio",
      "command": "fridare"
    }
  }
}

Gemini CLI

Add to ~/.gemini/settings.json:

{
  "mcpServers": {
    "fridare": {
      "command": "fridare",
      "args": []
    }
  }
}

Any MCP Client (Generic)

Fridare speaks stdio transport. To integrate with any MCP-compatible client:

command: fridare
transport: stdio

No arguments, no environment variables required. Just ensure fridare is on your PATH (i.e., pip install -e . was run in the active Python environment).

Tools

Device & Process

Tool	Description
`frida_devices`	List connected devices (USB/remote)
`frida_ps`	List processes with optional name filter
`frida_restart_server`	Kill & restart frida-server on device

Session

Tool	Description
`frida_attach`	Attach to process by name/PID (persists)
`frida_spawn`	Cold-start app + attach
`frida_detach`	Detach and cleanup
`frida_status`	Show session state (alive, PID, scripts, PCAP)

Execution

Tool	Description
`frida_exec`	Run arbitrary JS (supports `keep_previous` and `wait`)
`frida_rpc`	Call RPC exports on loaded script
`frida_load_script`	Load JS script from file
`frida_messages`	Read accumulated `send()` messages

Java Introspection

Tool	Description
`frida_list_classes`	Enumerate loaded classes (with filter)
`frida_list_methods`	List methods (supports `include_inherited`)
`frida_list_fields`	List fields of a class
`frida_hook`	Hook method — args, retval, optional backtrace

Network & Security

Tool	Description
`frida_ssl_capture`	Capture SSL + HTTP traffic (r0capture)
`frida_ssl_unpin`	Universal SSL pinning bypass
`frida_cert_dump`	Dump client certs + detect pinning
`frida_pcap_start`	Start saving traffic to PCAP file
`frida_pcap_stop`	Stop PCAP capture

Native

Tool	Description
`frida_list_modules`	List loaded .so modules + base addr
`frida_list_exports`	List native exports of a module

Usage

> Attach to Damai app
→ frida_attach("大麦")

> Search Mtop classes
→ frida_list_classes("Mtop")

> Hook setRetCode with stack trace
→ frida_hook("mtopsdk.mtop.domain.MtopResponse", "setRetCode", backtrace=true)

> Capture SSL traffic for 10s and save to pcap
→ frida_pcap_start("capture.pcap")
→ frida_ssl_capture(10)
→ frida_pcap_stop()

> Execute custom JS
→ frida_exec("Java.perform(function() { send({result: 'hello'}); })")

> Read hook output
→ frida_messages()

Architecture

fridare/
├── server.py          MCP tool definitions (22 tools)
├── session.py         Device/process/session lifecycle + PCAP writer
├── builtins.py        Built-in ops (class enum, hooking, SSL capture)
└── scripts/
    ├── ssl_capture.js  Universal traffic capture (SSL + HTTP + stack traces)
    ├── ssl_unpin.js    SSL pinning bypass
    └── cert_dump.js    Certificate dump + pinning locator

Design

System Frida — uses whatever frida version is installed. No bundled binaries, no version conflicts
Session persistence — one attach, unlimited operations. No reconnecting between commands
Message accumulation — hook output buffers in memory, read anytime with frida_messages()
PCAP export — frida_pcap_start/stop wraps traffic into Wireshark-compatible pcap files
Cross-platform — handles Windows/Linux/Mac quoting differences for adb commands

Credits

SSL capture based on r0capture by r0ysue
Built on Frida by Ole Andre V. Ravnas
MCP protocol by Anthropic

功能

持久会话 — attach 一次，无限执行
22 个 MCP 工具 — 在任何 MCP 兼容客户端中完整操控 Frida
r0capture 集成 — 通杀 SSL/HTTP 抓包 + PCAP 导出
SSL Pinning 绕过 — OkHttp / Conscrypt / WebView 通杀
证书导出 — 自动 dump 客户端证书为 P12
跨平台 — Windows / Linux / macOS，使用系统已装的 Frida，无版本冲突

安装

需要先在当前 Python 环境中准备好 frida 和 frida-tools。

从 GitHub 安装

pip install git+https://github.com/MyuriKanao/fridare.git

或者安装成独立命令行工具：

uv tool install git+https://github.com/MyuriKanao/fridare.git

本地开发安装

pip install -e .

发布

维护者可使用标准 Python 打包流程发布：

python -m pip install --upgrade build twine
python -m build
python -m twine check dist/*
python -m twine upload dist/*

如果想先验证发布流程，建议先上传到 TestPyPI，再发布到正式 PyPI。

GitHub Actions 自动发布

仓库已包含 .github/workflows/pypi-publish.yml：

PR / 推送到 main 时自动构建并执行 twine check
推送符合 v* 的 tag 时自动发布到 PyPI

典型发布流程：

git tag v0.1.0
git push origin v0.1.0

首次自动发布前，需要在 PyPI 中配置 Trusted Publisher：

PyPI 项目名：fridare
GitHub Owner：MyuriKanao
仓库名：fridare
Workflow 文件：.github/workflows/pypi-publish.yml
Environment：pypi

如果 PyPI 项目还不存在，先创建 pending publisher，再用第一次带 tag 的发布自动创建项目。

配置

Fridare 是标准 stdio MCP 服务器，可接入任何 MCP 兼容客户端：

Claude Code (CLI / 桌面 / Web)

claude mcp add fridare -- fridare

或手动编辑 ~/.claude.json：

{
  "mcpServers": {
    "fridare": {
      "type": "stdio",
      "command": "fridare"
    }
  }
}

Claude Desktop

Settings > Developer > Edit Config，编辑 claude_desktop_config.json：

{
  "mcpServers": {
    "fridare": {
      "command": "fridare"
    }
  }
}

Cursor

项目根目录 .cursor/mcp.json：

{
  "mcpServers": {
    "fridare": {
      "command": "fridare"
    }
  }
}

VS Code (GitHub Copilot)

.vscode/mcp.json：

{
  "servers": {
    "fridare": {
      "type": "stdio",
      "command": "fridare"
    }
  }
}

Windsurf

~/.codeium/windsurf/mcp_config.json：

{
  "mcpServers": {
    "fridare": {
      "command": "fridare"
    }
  }
}

Codex CLI (OpenAI)

~/.codex/config.json：

{
  "mcpServers": {
    "fridare": {
      "type": "stdio",
      "command": "fridare"
    }
  }
}

Gemini CLI

~/.gemini/settings.json：

{
  "mcpServers": {
    "fridare": {
      "command": "fridare",
      "args": []
    }
  }
}

通用 MCP 客户端

Fridare 使用 stdio 传输协议，无需额外参数或环境变量。确保 fridare 在 PATH 中即可（即在当前 Python 环境中执行过 pip install -e .）。

使用示例

你: attach 大麦 APP
→ frida_attach("大麦")

你: 搜 Mtop 相关的类
→ frida_list_classes("Mtop")

你: hook setRetCode，带堆栈
→ frida_hook("mtopsdk.mtop.domain.MtopResponse", "setRetCode", backtrace=true)

你: 抓 10 秒流量并存 pcap
→ frida_pcap_start("capture.pcap")
→ frida_ssl_capture(10)
→ frida_pcap_stop()

你: 拿 hook 产生的消息
→ frida_messages()

Name		Name	Last commit message	Last commit date
.github/workflows		.github/workflows
fridare		fridare
.gitignore		.gitignore
LICENSE		LICENSE
README.md		README.md
pyproject.toml		pyproject.toml

Latest commit

History

Folders and files

Repository files navigation

Fridare

Features

Install

From GitHub

For local development

Publishing

GitHub Actions release flow

Configure

Claude Code (CLI / Desktop / Web)

Claude Desktop

Cursor

VS Code (GitHub Copilot)

Windsurf

Codex CLI (OpenAI)

Gemini CLI

Any MCP Client (Generic)

Tools

Device & Process

Session

Execution

Java Introspection

Network & Security

Native

Usage

Architecture

Design

Credits

功能

安装

从 GitHub 安装

本地开发安装

发布

GitHub Actions 自动发布

配置

Claude Code (CLI / 桌面 / Web)

Claude Desktop

Cursor

VS Code (GitHub Copilot)

Windsurf

Codex CLI (OpenAI)

Gemini CLI

通用 MCP 客户端

使用示例

About

Resources

Stars

Watchers

Forks

Releases

Packages

Contributors

Languages