test-driven-development
Pass
Audited by Gen Agent Trust Hub on Feb 17, 2026
Risk Level: SAFE
Full Analysis
- Prompt Injection (SAFE): The instructions contain strong behavioral constraints ("Delete means delete") intended to enforce TDD methodology. These do not attempt to bypass safety filters or target the system prompt.
- Command Execution (SAFE): The skill utilizes standard local development commands (e.g.,
npm test) for verifying test results. No suspicious or high-privilege commands are present. - Data Exposure & Exfiltration (SAFE): No hardcoded credentials, sensitive file access, or unauthorized network operations were identified.
- Remote Code Execution (SAFE): No remote code execution patterns, external downloads, or untrusted script execution were found.
Audit Metadata