qa-planner
Pass
Audited by Gen Agent Trust Hub on Jul 3, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill explicitly mandates the redaction of sensitive data—including secrets, credentials, API keys, and PII—before any synchronization with external platforms such as Notion or Plane.so (Evidence: references/notion-mcp.md, references/plane-hybrid.md).
- [SAFE]: The integration with external platforms (Notion, Plane.so) is achieved through established MCP (Model Context Protocol) tools, which are standard for platform-integrated agent operations and do not involve unauthorized network requests.
- [SAFE]: The binary template provided (
Template Test Case.xlsx) follows standard Excel file structures and does not contain malicious macros or hidden executable code. - [SAFE]: While the skill processes untrusted external data (PRDs, user stories), it implements multiple safety gates, such as requirement gap checks, ambiguity resolution through user questioning, and human review loops, which mitigate the risk of indirect prompt injection.
Audit Metadata