Secrets Hygiene

Pass

Audited by Gen Agent Trust Hub on Aug 7, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill serves an educational and procedural purpose, offering best practices for incident response related to leaked credentials.
  • [CREDENTIALS_UNSAFE]: While the skill mentions credential formats and storage locations (such as .env files, ~/.ssh/id_rsa, and AWS/Stripe keys), it does so in a safe, educational context for detection purposes. It explicitly hardcodes only placeholder values (e.g., sk_live_REPLACE-WITH-REAL-KEY) and correctly advises users to manage secrets using dedicated managers or gitignored files.
  • [EXTERNAL_DOWNLOADS]: The skill recommends well-known security tools such as trufflehog, gitleaks, and detect-secrets. It does not provide direct download links or automated installation scripts that would bypass user review.
  • [PROMPT_INJECTION]: No attempts to override agent behavior or bypass safety filters were detected. The language is instructional and focused on security compliance.
  • [DATA_EXFILTRATION]: There are no network operations, subprocess calls, or patterns suggesting the unauthorized collection or transmission of data.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 7, 2026, 11:39 AM
Security Audit — agent-trust-hub — Secrets Hygiene