Vouch-agent is an open-source command-line security scanner for AI-agent skills. It performs deterministic static analysis of SKILL.md files for malware, prompt injection, and secret exfiltration, with an optional LLM analysis layer for Claude, Cursor, and MCP skills.
vouch-agent is an AI & LLM security project. It focuses on auditing AI-agent skills for malicious code, prompt injection, and secret exfiltration before use. vouch-agent is an open-source project aimed at AI-agent developers and security engineers. The project is open source (MIT). vouch-agent is available on the command line, and it can be self-hosted.
Wael Abouceo builds and maintains vouch-agent, and it first shipped in 2026. The project is developed in the open on GitHub with 57 commits in the last 90 days. Among its 7 catalogued features are static analysis, malware detection, and prompt injection detection. It exposes integrations via an MCP server.
Summary written by a language model from the project’s public pages.
What PulseGate has recorded for this listing
Closest matches by what these projects do