e2e-flow-verification
Pass
Audited by Gen Agent Trust Hub on Aug 4, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill contains markdown-based instructions for manual and automated frontend verification. It does not contain executable code, obfuscation, or remote resource fetching.
- [CREDENTIALS_UNSAFE]: No hardcoded credentials were found. Section 7 explicitly instructs the agent to 'Alert the human immediately' if it encounters credentials, secrets, or authentication walls, which is a security best practice.
- [DATA_EXFILTRATION]: No network exfiltration or sensitive data access patterns were identified. The skill focuses on testing local or development environments through tools like Playwright.
- [PROMPT_INJECTION]: The instructions are process-oriented and do not attempt to bypass safety filters or override system prompts. The language used ('Never declare work complete', 'Stop and say so') defines the operational flow for testing rather than attempting to subvert the agent's core instructions.
Audit Metadata