pre-flight
Pass
Audited by Gen Agent Trust Hub on Aug 23, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data via git commands. 1. Ingestion points: Git diff output in Phase 2. 2. Boundary markers: Absent. 3. Capability inventory: Repository inspection via git. 4. Sanitization: Absent.
Audit Metadata