Installation Guide¶

This guide covers the complete installation process for JADX-AI-MCP.

Prerequisites¶

System Requirements¶

Operating System: Windows, macOS, or Linux
Java: Version 11 or higher
Python: Version 3.10 or higher
JADX: Version 1.5.1+ (r2333+)

Check Prerequisites¶

# Check Java version
java -version

# Check Python version
python --version

# Check JADX version
jadx --version

Installation Methods¶

Method 1: One-Line Installation (Recommended)¶

The easiest way to install the plugin:

jadx plugins --install "github:zinja-coder:jadx-ai-mcp"

This command automatically downloads and installs the latest version of the plugin.

Method 2: Manual Installation via JADX-GUI¶

Download the latest release from GitHub Releases
Download both:
jadx-ai-mcp-<version>.jar
jadx-mcp-server-<version>.zip
Open JADX-GUI
Navigate to: Plugins → Install Plugin
Select the downloaded .jar file
Restart JADX-GUI

Plugin Installation

Method 3: Build from Source¶

# Clone the repository
git clone https://github.com/zinja-coder/jadx-ai-mcp.git
cd jadx-ai-mcp

# Build the plugin
./gradlew build

# The built JAR will be in build/libs/
# Install it using Method 2 above

MCP Server Setup¶

Step 1: Extract Server Files¶

# Download server package
wget https://github.com/zinja-coder/jadx-ai-mcp/releases/latest/download/jadx-mcp-server.zip

# Extract
unzip jadx-mcp-server.zip
cd jadx-mcp-server

Step 2: Install UV Package Manager¶

UV is a fast Python package manager used by this project:

# Linux/macOS
curl -LsSf https://astral.sh/uv/install.sh | sh

# Windows (PowerShell)
powershell -c "irm https://astral.sh/uv/install.ps1 | iex"

Step 3: Verify Installation¶

# Check UV installation
uv --version

# Test server (should show usage info)
uv run jadx_mcp_server.py --help

Optional: Create Virtual Environment¶

# Create venv (optional but recommended for troubleshooting)
uv venv
source .venv/bin/activate  # Linux/macOS
# OR
.venv\Scripts\activate  # Windows

# Install dependencies manually
uv pip install httpx fastmcp

LLM Client Configuration¶

Claude Desktop¶

Linux Configuration¶

# Edit config file
nano ~/.config/Claude/claude_desktop_config.json

Windows Configuration¶

Location: %APPDATA%\Claude\claude_desktop_config.json

macOS Configuration¶

Location: ~/Library/Application Support/Claude/claude_desktop_config.json

Configuration Content¶

{
  "mcpServers": {
    "jadx-mcp-server": {
      "command": "/path/to/uv",
      "args": [
        "--directory",
        "/path/to/jadx-mcp-server/",
        "run",
        "jadx_mcp_server.py"
      ]
    }
  }
}

Important: Replace paths with absolute paths on your system.

Find UV Path¶

# Linux/macOS
which uv

# Windows
where uv

Alternative: Install as UV Tool¶

# Install globally as a UV tool
uv tool install git+https://github.com/zinja-coder/jadx-mcp-server

# Then use simplified config
{
  "mcpServers": {
    "jadx-mcp-server": {
      "command": "jadx_mcp_server"
    }
  }
}

Cherry Studio Configuration¶

Open Cherry Studio settings
Navigate to MCP configuration
Add new MCP server:
Type: stdio
Command: uv

Arguments:

--directory
/path/to/jadx-mcp-server
run
jadx_mcp_server.py

LM Studio Configuration¶

Open LM Studio
Navigate to MCP settings
Edit mcp.json:

{
  "mcpServers": {
    "jadx-mcp-server": {
      "command": "/path/to/uv",
      "args": [
        "--directory",
        "/path/to/jadx-mcp-server",
        "run",
        "jadx_mcp_server.py"
      ]
    }
  }
}

Remote & Custom Configuration¶

CLI Flags Reference¶

There are two separate connections and each has its own host/port.

┌─────────────┐    --host / --port     ┌──────────────────┐   --jadx-host / --jadx-port   ┌──────────────────┐
│  LLM Client │ ◄──────────────────►   │  jadx-mcp-server │ ──────────────────────────►   │  JADX-GUI Plugin │
│  (Claude,   │   Where the MCP server │                  │   Where the MCP server looks  │  (jadx-ai-mcp)   │
│   Codex..)  │   LISTENS for clients  │                  │   for the JADX plugin         │                  │
└─────────────┘                        └──────────────────┘                               └──────────────────┘

Flag	Default	Controls
`--http`	off	Use HTTP transport instead of stdio
`--host`	`127.0.0.1`	Where the MCP server listens (bind address for LLM clients)
`--port`	`8651`	Which port the MCP server listens on
`--jadx-host`	`127.0.0.1`	Where to find the JADX plugin (the target JADX-GUI machine)
`--jadx-port`	`8650`	Which port the JADX plugin is on

HTTP Stream Mode (Optional)¶

Run the server in HTTP mode for remote access from an LLM client:

# Default HTTP mode (localhost:8651)
uv run jadx_mcp_server.py --http

# Custom port
uv run jadx_mcp_server.py --http --port 9999

Remote / Docker / WSL Access¶

To make the MCP server accessible from other machines (e.g., if you run the MCP server in Docker, or via WSL, and the LLM client is on the host Windows machine):

# Bind to all interfaces
uv run jadx_mcp_server.py --http --host 0.0.0.0

# Bind to all interfaces on a custom port
uv run jadx_mcp_server.py --http --host 0.0.0.0 --port 9999

Security Warning — Remote Binding

When using --host 0.0.0.0 (or any non-localhost address), the MCP server binds to all network interfaces over plain HTTP with no authentication. - Anyone on the network can connect and invoke all MCP tools. - There is no TLS encryption — traffic can be intercepted. - An attacker can use the server to read decompiled code, modify it, and access debug info. - Mitigation: Only bind to 0.0.0.0 on trusted, isolated networks (e.g., Docker bridge), use a firewall, or tunnel via SSH.

Remote JADX Plugin Configuration¶

If the JADX AI MCP Plugin is running on a different machine (e.g., JADX on a remote VM, MCP server on your local host), use the --jadx-host option:

# Connect to JADX plugin on a remote host
uv run jadx_mcp_server.py --jadx-host 192.168.1.100 --jadx-port 8650

Custom Plugin Port Configuration¶

If you change the JADX-GUI Plugin port:

Open JADX-GUI with the plugin installed
Navigate to: Plugins → JADX-AI-MCP → Configure Port
Enter desired port (default: 8650)
Click Restart Server

Port Configuration

When using custom plugin port:

# Connect to plugin on custom port
uv run jadx_mcp_server.py --jadx-port 8652

Update LLM client configuration appropriately:

{
  "mcpServers": {
    "jadx-mcp-server": {
      "command": "/path/to/uv",
      "args": [
        "--directory",
        "/path/to/jadx-mcp-server/",
        "run",
        "jadx_mcp_server.py",
        "--jadx-host",
        "192.168.1.100",
        "--jadx-port",
        "8652"
      ]
    }
  }
}

Verification¶

1. Check Plugin Status¶

Open JADX-GUI
Load any APK file
Check: Plugins → JADX-AI-MCP → Server Status
Should show: "Server Running on port 8650"

2. Check MCP Server Connection¶

Open your LLM client (Claude/Cherry Studio/LM Studio)
Look for hammer icon (🔨) or MCP tools indicator
Click and verify "jadx-mcp-server" is listed
Should show ~30 available tools

3. Test Basic Functionality¶

Run this prompt in your LLM client:

List all available JADX MCP tools

Expected response: List of 30+ tools including fetch_current_class, get_android_manifest, etc.