triage

Pass

Audited by Gen Agent Trust Hub on Jul 10, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [PROMPT_INJECTION]: The skill processes untrusted input from external issue trackers that could contain malicious instructions designed to influence agent behavior.
  • Ingestion points: The agent reads the full issue body, comments, and reporter metadata during the triage process in SKILL.md.
  • Boundary markers: There are no explicit delimiters or specific instructions to the agent to disregard instructions embedded within the issue data.
  • Capability inventory: The skill is capable of modifying issue labels, posting comments, closing issues, and writing files to the local .out-of-scope/ directory.
  • Sanitization: Content retrieved from the issue tracker is not validated or sanitized before being incorporated into the agent's context.
  • Contextual Mitigation: The instructions mandate that the agent recommend actions and wait for maintainer direction or confirmation before proceeding with transitions or closures, providing a human-in-the-loop control point.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 10, 2026, 06:43 AM
Security Audit — agent-trust-hub — triage