MCP Catalogs
Homejadx-ai-mcp screenshot

jadx-ai-mcp

by zinja-coder·2,023·Score 54

MCP server for JADX decompiler that enables AI-powered APK analysis with tools for reverse engineering and vulnerability detection.

securitydeveloper-toolsai-llm
197
Forks
10
Open issues
1 mo ago
Last commit
2d ago
Indexed

Overview

JADX-AI-MCP is a comprehensive plugin for the JADX decompiler that integrates Model Context Protocol (MCP) to enable real-time reverse engineering and static code analysis. It allows LLMs like Claude to interact with decompiled Android APKs through various MCP tools that can fetch class source, analyze manifest files, search methods, and even perform debugging operations. The server consists of two main components: the JADX-AI-MCP Java plugin and a standalone Python MCP server that facilitates communication with the LLM client.

Try asking AI

After installing, here are 5 things you can ask your AI assistant:

you:AI-assisted reverse engineering of Android APKs to understand application structure and behavior
you:Automated vulnerability detection in Android applications through AI analysis of decompiled code
you:Real-time code review and refactoring suggestions during debugging sessions
you:What LLM clients are compatible with JADX-AI-MCP?
you:Can I use JADX-AI-MCP without JADX GUI?

When to choose this

Choose JADX-AI-MCP when you need to analyze Android APKs using AI assistance without leaving your decompiler environment, especially for reverse engineering and vulnerability detection tasks.

When NOT to choose this

Avoid if you need to analyze non-Android applications, as this tool is specifically designed for APK analysis only.

Tools this server exposes

12 tools extracted from the README
  • fetch_current_class

    Get the class name and full source of selected class

  • get_selected_text

    Get currently selected text

  • get_all_classes

    List all classes in the project

  • get_class_source

    Get full source of a given class

  • search_classes_by_keyword

    Search for classes containing a specific keyword

  • get_android_manifest

    Retrieve the AndroidManifest.xml content

  • rename_class

    Rename a class in the decompiler

  • xrefs_to_class

    Find all references to a class

  • get_strings

    Fetch the strings.xml file

  • get_main_activity_class

    Fetch the main activity class from AndroidManifest.xml

  • debug_get_stack_frames

    Get stack frames from the JADX debugger

  • debug_get_variables

    Get variables from the JADX debugger

Comparable tools

apktool-mcp-serverghidra-mcpmobSF-mcpandroguard-python

Installation

Installation

  1. Download the latest releases from [GitHub](https://github.com/zinja-coder/jadx-ai-mcp/releases)
  2. Extract the jadx-mcp-server-<version>.zip file
  3. Install Python dependencies: pip install -r jadx-mcp-server/requirements.txt
  4. Run JADX with the plugin: java -jar jadx-ai-mcp-<version>.jar
  5. Start the MCP server: python jadx_mcp.py

Claude Desktop configuration:

{
  "mcpServers": {
    "jadx": {
      "command": "python",
      "args": ["/path/to/jadx_mcp.py"]
    }
  }
}

FAQ

What LLM clients are compatible with JADX-AI-MCP?
The MCP server is designed to work with any LLM client that supports the Model Context Protocol, including Claude Desktop, OpenAI ChatGPT, and other compatible clients.
Can I use JADX-AI-MCP without JADX GUI?
Currently, JADX-AI-MCP requires JADX GUI to function as it depends on the GUI components to perform actions and gather data from the decompiled application.

Compare jadx-ai-mcp with

GitHub →

Last updated · Auto-generated from public README + GitHub signals.