Radare2 MCP Server
by dnakov
Radare2 MCP Server enables the use of radare2 with AI assistants like Claude. It provides a simple MCP server for seamless binary analysis and file exploration.
Last updated: N/A
What is Radare2 MCP Server?
Radare2 MCP Server is a server that allows you to use the radare2 reverse engineering framework with AI assistants like Claude. It acts as a bridge, enabling the AI to interact with radare2's powerful analysis capabilities.
How to use Radare2 MCP Server?
To use the server, install it using r2pm. Then, configure your AI assistant (e.g., Claude Desktop) to use the server by adding the appropriate configuration to the AI's configuration file, specifying the command and arguments to launch the r2mcp executable or Docker container.
Key features of Radare2 MCP Server
Direct stdin/stdout communication model
Basic tool capabilities
Seamless binary analysis with radare2
Direct integration of radare2 with AI assistants
File exploration and inspection
Use cases of Radare2 MCP Server
Automated binary analysis with AI assistance
Vulnerability discovery
Malware analysis
Reverse engineering tasks
FAQ from Radare2 MCP Server
What is MCP?
What is MCP?
MCP stands for Message Channel Protocol, a protocol used for communication between applications.
How do I install r2mcp?
How do I install r2mcp?
The simplest way to install r2mcp is by using the command r2pm -Uci r2mcp.
Where is the Claude Desktop configuration file located?
Where is the Claude Desktop configuration file located?
The location varies depending on your operating system: macOS: ~/Library/Application Support/Claude/claude_desktop_config.json, Windows: %APPDATA%\Claude\claude_desktop_config.json, Linux: ~/.config/Claude/claude_desktop_config.json.
Can I use Docker to run r2mcp?
Can I use Docker to run r2mcp?
Yes, you can use Docker. The README provides instructions on how to build and run the r2mcp Docker image.
Why can't I run r2mcp directly from the shell?
Why can't I run r2mcp directly from the shell?
The r2mcp executable is designed to be launched by the MCP service handler of your language model and not directly from the shell.