setup

Pass

Audited by Gen Agent Trust Hub on Nov 22, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill explicitly instructs the agent to keep credentials out of the generated configuration file (ai/config/tickets.md), promoting secure secret management practices.
  • [SAFE]: A human-in-the-loop requirement is enforced, where the agent must present proposed settings for approval before saving them, mitigating risks from automated actions.
  • [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted repository data (agent instructions and conventions) to derive its configuration.
  • Ingestion points: SKILL.md (inspects repository, Git remote, agent instructions, ai/ conventions).
  • Boundary markers: Absent.
  • Capability inventory: File system write (ai/config/tickets.md).
  • Sanitization: Absent, but mitigated by mandatory human approval of proposed settings.
Audit Metadata
Risk Level
SAFE
Analyzed
Nov 22, 2026, 11:06 AM