ext-triage-reqs

Pass

Audited by Gen Agent Trust Hub on Jul 11, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [PROMPT_INJECTION]: The skill presents an indirect prompt injection surface by processing and acting upon automated PR review comments from external AI tools and linters.
  • Ingestion points: The logic in 'standards/pr-comment-disposition.md' ingests comments from 'gemini-code-assist', 'Copilot', 'AsciiDoc-lint', and other external tools.
  • Boundary markers: The skill relies on a 'validity check' and 'plan-intent check' performed by a separate agent ('persona-plan-marshall-agent') to ensure suggestions do not contradict the project plan.
  • Capability inventory: The triage disposition can lead to follow-up commits ('FIX'), automated replies, and thread resolution.
  • Sanitization: There is no documentation of sanitization, escaping, or strict schema validation for the ingested comment text before the agent determines the disposition.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 11, 2026, 07:13 AM
Security Audit — agent-trust-hub — ext-triage-reqs