LaunchKit · 2026
Back to Skills

agents-skill-security-audit

Minimal helper to audit skill.md-style instructions for supply-chain risks.

0
683 downloads
by @cerbug45

Setup & Installation

openclaw skills install @cerbug45/agents-skill-security-audit

Or with the ClawHub CLI, for registry-managed skill folders outside a full OpenClaw workspace:

npx clawhub install agents-skill-security-audit

What This Skill Does

Scans skill.md-style instruction files for supply-chain risks using heuristic pattern matching. It flags exfiltration patterns such as HTTP POST requests, curl calls to unknown domains, reads of sensitive files like ~/.env, and credential keywords. Output is a markdown report with a risk level.

It targets the specific structure of skill.md instruction files rather than running generic secrets scanning, so it catches context-aware patterns like credential reads embedded in agent prompts.

When to use it

  • Checking a third-party skill file before installing it
  • Auditing new agent instructions added to a shared team repo
  • Scanning downloaded SKILL.md files before running them locally
  • Adding a skill file check to a CI pipeline before merging
  • Spot-checking an AI agent prompt file for credential leaks

Example Workflow

Here's how your AI assistant might use this skill in practice.

INPUT

User asks: audit this skill.md for supply-chain risks

AGENT
  1. 1Receives the path to the skill.md file
  2. 2Runs audit.py against the file using heuristic pattern matching
  3. 3Scans for HTTP POST calls, curl to unknown domains, ~/.env reads, and credential keywords
  4. 4Builds a permission manifest listing filesystem and network touches found
  5. 5Outputs a markdown summary with an overall risk level
OUTPUT

Markdown report listing flagged patterns and an overall risk level rating