escalation-router
Pass
Audited by Gen Agent Trust Hub on Feb 21, 2026
Risk Level: SAFEPROMPT_INJECTIONNO_CODE
Full Analysis
- [Indirect Prompt Injection] (LOW): The skill processes untrusted data from an external source (fix-engine error reports). Ingestion points: The 'error_report' variable within the research_request section. Boundary markers: Absent; the data is interpolated directly into a YAML-like structure without delimiters. Capability inventory: The skill signals status updates and requests file creation (./handoff-prd-{section-id}.md) via a sub-agent. Sanitization: No input validation or content escaping is specified.
- [Data Exposure] (SAFE): No hardcoded secrets, API keys, or sensitive system file paths were detected.
Audit Metadata