🛠 burpwn: Web Pentesting with AI Agents
burpwn has been released — a tool for conducting web pentesting using AI agents. It functions as a transparent proxy, a sandbox, and an interface adapted for programmatic agent control. The system utilizes a rootless Linux sandbox and TLS-MITM to decrypt traffic, allowing agents to analyze requests within an isolated environment.
🌍 It addresses security concerns when using autonomous AI agents. It enables automated auditing while ensuring process isolation from the host system and providing detailed monitoring of network traffic.
👤 Security researchers and AI agent developers gain a tool for the safe and controlled "probing" of web applications via CLI or an MCP server.
Source 1: https://github.com/own2pwn-fr/burpwn