figure-it-out
Pass
Audited by Gen Agent Trust Hub on Jul 10, 2026
Risk Level: SAFECOMMAND_EXECUTIONREMOTE_CODE_EXECUTIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The instructions direct the agent to utilize host terminal commands, git tools, and browser access to perform migrations and code changes. This is consistent with the skill's purpose as a development aid.
- [REMOTE_CODE_EXECUTION]: The skill instructs the agent to launch sub-processes and CLI workers using 'spawn-codex-worker' and 'codex exec'. This represents a mechanism for task delegation to subagents within the expected execution environment.
- [PROMPT_INJECTION]: The skill contains instructions to override default agent behaviors, specifically directing the agent to use particular models and host-appropriate tools over panel defaults. These are functional behavioral overrides used to define the agent's logic for complex tasks rather than attempts to bypass safety filters.
Audit Metadata