ds-optimize
Pass
Audited by Gen Agent Trust Hub on Jun 20, 2026
Risk Level: SAFECOMMAND_EXECUTION
Full Analysis
- [COMMAND_EXECUTION]: The skill mandates a strict execution discipline where all terminal-based operations (Git, Python, package management, file inspection) must be performed exclusively through the
bash_exectool. - [SAFE]: The instructions implement robust state management protocols, requiring the agent to recover the optimization frontier and recent memory before proposing or executing new tasks.
- [SAFE]: The skill enforces a defensive 'smoke-before-long-run' policy and requires the agent to analyze logs and identify root causes before retrying failed operations.
- [SAFE]: The workflow incorporates rigorous tracking through mandatory local control files (
OPTIMIZE_CHECKLIST.mdandCANDIDATE_BOARD.md) which document decisions, candidates, and experimental results. - [SAFE]: The instructions are designed to prevent accidental or redundant execution by requiring the inspection of existing managed sessions before launching new ones.
Audit Metadata