autopilot
Pass
Audited by Gen Agent Trust Hub on Sep 16, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill defines a structured workflow involving three phases:
deep-interview,ralplan, andultragoal. It manages internal state via a session-scoped JSON file (autopilot-state.json). - [SAFE]: No external network operations, remote script downloads, or package installations are identified. All operations are local to the agent's lifecycle.
- [SAFE]: The execution policy and state management sections focus on internal logic, error recovery, and persistence of task-related artifacts rather than sensitive user data or system credentials.
- [SAFE]: There is no evidence of obfuscation, privilege escalation, or malicious prompt injection patterns. The use of
$commandstyle markers refers to internal skill invocations or aliases within the orchestrator framework.
Audit Metadata