triage-helper

Pass

Audited by Gen Agent Trust Hub on Jun 17, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill instructions and metadata do not contain any malicious patterns or dangerous operations.
  • [PROMPT_INJECTION]: Indirect Prompt Injection surface analysis:
  • Ingestion points: The skill retrieves external data from the official flutter/flutter-intellij GitHub repository (SKILL.md).
  • Boundary markers: Instructions do not specify explicit delimiters for issue content.
  • Capability inventory: The agent is restricted to read operations and presenting suggestions; it has no file-write or subprocess capabilities defined in this skill.
  • Sanitization: No explicit sanitization of issue content is mentioned.
  • Note: The risk of indirect prompt injection is mitigated by the explicit instruction to avoid direct actions (labeling, commenting, closing) and to wait for user feedback for every step, maintaining a human-in-the-loop.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 17, 2026, 11:37 AM
Security Audit — agent-trust-hub — triage-helper