qa-reporter
Pass
Audited by Gen Agent Trust Hub on Jul 3, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill follows least-privilege and high-governance principles. It mandates that no issues be submitted or states changed without explicit user approval and verified configuration.
- [SAFE]: It implements an 'Approval Blocking-Info Guard' in SKILL.md, which programmatically checks for missing context (e.g., environment, test data, OTP) and blocks automated approval until information is provided or a second confirmation is given.
- [SAFE]: Integration with Plane and Notion is handled via pre-configured MCP tools. The skill includes specific instructions to redact credentials and PII, ensuring that sensitive data is not leaked into shared reports or comments.
- [SAFE]: The skill uses a 'State Router' that forces the agent to verify state transitions in the external tracker (Plane) via read-back checks before proceeding, mitigating risks of desynchronization or unintended automated actions.
- [SAFE]: Destructive actions like archiving or deleting work items require explicit user intent and a second confirmation naming the specific identifiers, preventing accidental data loss.
Audit Metadata