seppuku
Pass
Audited by Gen Agent Trust Hub on Aug 26, 2026
Risk Level: SAFENO_CODE
Full Analysis
- [SAFE]: The skill implements a self-termination procedure for conversation threads. It provides structured instructions for stopping work, creating a handoff summary, and archiving the session.
- [SAFE]: Includes specific security constraints that forbid the agent from deleting files, sessions, logs, or mutating project state (such as GitHub) without separate, explicit user approval.
- [SAFE]: Contains guardrails to prevent accidental or malicious triggering by instructing the agent not to infer termination intent from errors or frustration, requiring instead a clear and explicit user request.
- [SAFE]: No executable code, network operations, or external dependencies are included in this skill.
Audit Metadata