skill-security
superagent-aiKnow whether an agent skill is safe to install.
agent skill security
What it does
Determines whether an AI agent skill is safe to trust by combining static findings with a behavior-versus-description review.
When to use it
Use when the user wants to audit the safety of a skill, plugin, SKILL.md, or agent tool.
How to use it
Give it a skill folder, manifest, archive, or repository; it scans and reviews the contents and returns a cited security audit.
What you provide
- a skill folder, file, archive, or repo
Access · 4
This skill
referenced skill repository
Read
Fetches remote skill repositories
local path
Write
Clones skills to a local path
target skill files
Read
Reads the target skill files
temporary directory
Write
Extracts supplied archives temporarily
What you need · 1
Python 3 is required to run the bundled scanner.
About this skill
- Visibility
- Public
- Repository
- superagent-ai/skills
- Created
- Oct 8, 2026
- Updated
- Oct 8, 2026
- Files
- 8