ai-regression-testing
Pass
Audited by Gen Agent Trust Hub on Apr 11, 2026
Risk Level: SAFE
Full Analysis
- [COMMAND_EXECUTION]: The skill provides templates for a bug-check workflow that involves executing standard build and test commands (
npm run test,npm run build). These commands are standard development practices and are not used maliciously in this context.- [DATA_EXFILTRATION]: All network-related code snippets focus on local testing (http://localhost:3000). No exfiltration patterns to external domains or sensitive data exposure were identified.- [PROMPT_INJECTION]: The skill uses strong instructional language ("mandatory, cannot skip") within its command templates. However, these are localized to the suggested regression testing workflow and do not attempt to bypass core agent safety guidelines or hijack the global system prompt.- [SAFE]: Sensitive environment variables (e.g., Supabase keys) are provided as empty string placeholders in the example setup files, following security best practices for mock configurations.
Audit Metadata