Skip to content
Mike Reams

Try

Searches titles, summaries and topics across posts, work, diagrams, pages and the Toolbox resources.

↑ ↓ move · Enter open · Esc close
← AI Library

Skill · Curated · Security

Skill security auditor

Scans a skill before you install it for prompt injection, hidden text, network calls, shell execution and credential reads.

What you get

A pass or fail report on a skill folder, with each risky line flagged.

What I changed, and why I run it

  • Every skill on this page went through it, then a manual read of every file.
  • Use it as a first pass, not the only gate: its curl rule misses file uploads sent with curl -F.
  • supply chain
  • prompt injection
  • skills

Source

Author
Alireza Rezvani
Licence
MIT
Reviewed at
aa8d778

Reviewed 2026-08-20 Passed, with notes: every file read by hand and checked for prompt injection, hidden text, network calls, shell execution and credential reads. Second opinion: skills.sh audits (opens in a new tab).

Not hosted here: get it from the author, at the commit I reviewed.

Updated 2026-08-20