finalrun-update-test
Pass
Audited by Gen Agent Trust Hub on Feb 27, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill uses specific MCP tools (e.g., update_tests_by_name, delete_tests_by_name) to manage test prompts. All operations that modify state require a two-phase confirmation process involving a preview and a confirmation token.
- [SAFE]: The instructions explicitly forbid the agent from modifying source code or inventing sensitive data such as credentials, OTPs, or user IDs. The agent is required to pause and ask the user for these values if they are unknown.
- [SAFE]: A mandatory 'Review and Confirm' step is included in the workflow, ensuring the user must approve the update plan before any modifications are executed via the MCP tools.
- [SAFE]: No suspicious patterns such as prompt injection, obfuscation, or unauthorized data exfiltration were detected in the skill instructions.
Audit Metadata