ci-cd-and-automation
Pass
Audited by Gen Agent Trust Hub on Jun 16, 2026
Risk Level: SAFE
Full Analysis
- [PROMPT_INJECTION]: No prompt injection patterns or instructions to bypass safety guidelines were found. The instructional language used is benign and context-appropriate.
- [CREDENTIALS_UNSAFE]: No hardcoded secrets were detected. The skill correctly demonstrates using GitHub Secrets for sensitive values like database passwords and deployment tokens, and provides best-practice guidance on environment variable management.
- [EXTERNAL_DOWNLOADS]: Uses standard package managers (NPM) and well-known CLI tools (Playwright, Vercel) for CI tasks. These are routine operations within the context of build pipelines.
- [DATA_EXFILTRATION]: No evidence of unauthorized data access or network exfiltration patterns.
- [COMMAND_EXECUTION]: Contains standard development and build commands (linting, testing, building) expected for CI/CD automation. No dangerous or obfuscated command execution was found.
Audit Metadata