The Agent Skills Directory

[SAFE]: The skill establishes a structured process for creative work, emphasizing user approval through a 'HARD-GATE' mechanism before implementation actions occur. This process encourages safety and alignment by preventing unauthorized automated changes to the codebase.- [COMMAND_EXECUTION]: The skill utilizes git commands to commit design documentation. This behavior is restricted to the specific task of documenting design specs within the project's own repository and is consistent with standard development workflows.- [DATA_EXPOSURE]: The agent reads local project context, including files, documentation, and recent commit history. This is localized to the workspace and necessary for providing relevant design suggestions based on existing code patterns.

brainstorming