pwno-mcp
by pwno-io·★ 264·Score 49
MCP server that exposes GDB+pwndbg debugging capabilities for binary research and exploit development.
Overview
pwno-mcp is a stateful system designed for autonomous binary research and exploit development, specifically built to work with LLM agents. It runs GDB with pwndbg in an isolated environment, exposing debugging capabilities, exploit I/O, and helper tooling through the Model Context Protocol. The server supports deterministic execution control, fast context snapshots for registers, stack, disassembly, and backtrace, along with interactive exploit workflows through pwncli.
Try asking AI
After installing, here are 3 things you can ask your AI assistant:
When to choose this
Security researchers and reverse engineers who want LLM agents to assist with binary analysis and exploit development in an isolated environment.
When NOT to choose this
You shouldn't use this if you need commercial licensing without contacting the authors, or if you require architectures beyond what GDB+pwndbg supports.
Comparable tools
Installation
Installation
Using Docker (Recommended)
mkdir -p ./workspace
cp ./path/to/your/binary ./workspace/chal
chmod +x ./workspace/chaldocker run --rm -p 5500:5500 \
--cap-add=SYS_PTRACE \
--cap-add=SYS_ADMIN \
--security-opt seccomp=unconfined \
--security-opt apparmor=unconfined \
-v "$PWD/workspace:/workspace" \
ghcr.io/pwno-io/pwno-mcp:latestDefault MCP endpoint: http://127.0.0.1:5500/mcp
Claude Desktop Configuration
Add to Claude Desktop's config.json:
"mcpServers": {
"pwno-mcp": {
"command": "docker",
"args": ["run", "--rm", "-p", "5500:5500", "--cap-add=SYS_PTRACE", "--cap-add=SYS_ADMIN", "--security-opt", "seccomp=unconfined", "--security-opt", "apparmor=unconfined", "-v", "$PWD/workspace:/workspace", "ghcr.io/pwno-io/pwno-mcp:latest"]
}
}Compare pwno-mcp with
Last updated · Auto-generated from public README + GitHub signals.