Observed arrival · 2026-11-02
Agent Aisle audits agent skills before you trust them
A free, open-source scanner checks Agent Skills as text, grades them A to F, and lists audited skills for Claude Code, Codex, and Cursor.
○Open
⊠Login
$Paid
†Ads
✦Pretty
●Pro
◎Niche
◉Human
⚑Risk
ƒJS
- For
- Developers installing or publishing Agent Skills
- Worth noticing
- Community submissions are audited at a pinned commit, and a bot comments the grade on the pull request.
Field notes
The page describes checks for secrets, network calls, risky commands, hidden instructions, and unusual characters. Its example flags a script that sends environment variables to a remote endpoint, while a warning says automated checks can miss determined attempts to evade detection. The homepage reports 26 audited skills and shows a workflow for audits on folders, GitHub repositories, and community submissions.
Observed signals
Read the marks
Editorial observations of this landing page, not a rating.
○OpenPublic substance visible
✦PrettyNotable craft visible
●ProPolished or operationally mature
◎NicheUnusually specific use
ƒJavaScriptBrowser-side code central
One card from the complete issue