Launching today
Agent Aisle
A free, open-source security audit for Agent Skills
1 follower
A free, open-source security audit for Agent Skills
1 follower
A free, open-source security audit for Agent Skills. It reads a skill as text, never runs it, and grades it A to F. Plus a store of audited skills for Claude Code, Codex and Cursor.

Free
Launch Team

AWS Activate CreditsCredits that fuel your startupβs next milestone
Promoted
Hi Product Hunt! I'm Hazem, a young fullstack engineer.
AI coding agents can be extended with "skills": instructions and scripts the agent follows with your access to files and credentials. Most people install them without reading them, and researchers have already found hundreds of malicious ones this year.
Agent Aisle is my answer, in three parts:
π Audit: npx @agentsaisle/cli audit ./my-skill reads a skill as text, never runs it, checks 41 rules and grades it A to F with the exact file and line for every finding.
π Store: aisle add <skill> installs a skill pinned to an exact commit and audited first; aisle remove takes it out. Works with any GitHub skill, not only mine.
πΊοΈ arch-visualize: type /arch-visualize and get an offline web page of your system's architecture, with animated request flows and possible risks.
Everything is free and MIT licensed. I'd love your feedback, especially false positives from the audit or skills you'd like to see in the store. Thank you for checking it out!