skills/codiume/skills/pre-flight/Gen Agent Trust Hub

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
Risk Level
SAFE
Analyzed
Aug 23, 2026, 02:37 PM
Security Audit — agent-trust-hub — pre-flight