Auditing Skills

When you're publishing Claude skills on skills.sh or Tessl, this walks you through interpreting their security and quality audits, then fixing what they flag. It covers the three skills.sh scanners (Gen Agent Trust Hub, Socket, Snyk) and how to read their Pass/Warn/Fail verdicts on individual skill pages, not just repo listings. Most findings fall into a few buckets: credential handling (stop telling agents to paste tokens in config files), indirect prompt injection (don't let fetched API responses control agent behavior), and unverifiable dependencies (add provenance notes for your own tools, pin versions for everyone else's). The Tessl section decodes their activation and implementation scores, which mostly catch vague descriptions and bloated instructions. Includes reusable remediation templates so you can fix the same issue across twenty skills without rewriting the guidance each time.

GitHub stars: 553
last updated: May 19, 2026

View on GitHub

Install

$ npx -y skills add dbt-labs/dbt-agent-skills --skill auditing-skills --agent claude-code

Installs into .claude/skills of the current project.

Files

Auditing Skills

Audit published skills against third-party security scanners and quality reviewers, and remediate findings.

Security Audit Sources

skills.sh

skills.sh runs three independent security audits on every published skill:

Auditor	Focus	Detail Page Pattern
Gen Agent Trust Hub	Remote code execution, prompt injection, data exfiltration, command execution	`/security/agent-trust-hub`
Socket	Supply chain and dependency risks	`/security/socket`
Snyk	Credential handling, external dependencies, third-party content exposure	`/security/snyk`

Each auditor assigns one of: Pass, Warn, or Fail.

How to Check

Listing page — https://skills.sh/{org}/{repo} shows all skills but may not surface per-skill audit statuses
Individual skill pages — https://skills.sh/{org}/{repo}/{skill-name} shows the three audit badges (Pass/Warn/Fail)
Detailed findings — https://skills.sh/{org}/{repo}/{skill-name}/security/{auditor} where {auditor} is agent-trust-hub, socket, or snyk