brainstorming
Pass
Audited by Gen Agent Trust Hub on Feb 27, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill provides a structured methodology for turning ideas into technical designs through dialogue.
- [COMMAND_EXECUTION]: The skill instructs the agent to write documentation to the local filesystem (
docs/plans/) and perform Git operations (commits and worktree creation). These are standard actions for a developer-oriented tool and occur within the local project context. - [DATA_EXPOSURE]: The skill reads project files, documentation, and commit history to gather context. This is expected behavior for an agent designed to assist with project-specific design tasks.
- [PROMPT_INJECTION]: The description uses directive language ("You MUST use this") to ensure the agent follows the brainstorming process for complex tasks. This is a workflow instruction rather than an attempt to bypass safety filters or ignore system instructions.
Audit Metadata