mission-approved-doc
Pass
Audited by Gen Agent Trust Hub on May 27, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: No high-severity security issues were detected. The skill is designed with a strong security boundary through a mandatory approval process.
- [PROMPT_INJECTION]: The skill possesses an indirect prompt injection surface because it processes untrusted Markdown documents. However, this risk is mitigated by design.
- Ingestion points: The skill reads design documents and implementation plans from
docs/superpowers/specs/anddocs/superpowers/plans/as defined indoc-field-mapping.md. - Boundary markers: It implements a mandatory 'Phase 1: 批准门验证 (HARD STOP)' which strictly prevents execution until a user explicitly approves the document content.
- Capability inventory: The skill generates CSV files that define tasks requiring MCP tools such as
chrome-devtools,playwright, andscreenpipefor the subsequent execution stage. - Sanitization: The skill relies on human review as the primary validation mechanism; it maps extracted text to CSV fields following structured rules but does not perform automated content sanitization beyond the human-in-the-loop gate.
Audit Metadata