agent-skills-framework

Pass

Audited by Gen Agent Trust Hub on May 18, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill provides structured engineering workflows that include automated quality gates and verification steps to guide agent behavior.
  • [SAFE]: Installation procedures utilize official GitHub repositories and established platforms like Claude Code and Gemini CLI.
  • [SAFE]: Security guidelines correctly advise users to manage secrets via environment variables instead of hardcoding them in scripts or prompts.
  • [SAFE]: The framework acknowledges and mitigates indirect prompt injection risks by defining anti-rationalization tables and recommending strict input validation (e.g., using Zod or Joi).
Audit Metadata
Risk Level
SAFE
Analyzed
May 18, 2026, 08:09 PM
Security Audit — agent-trust-hub — agent-skills-framework