The Agent Skills Directory

[SAFE]: The skill consists entirely of instructional markdown that establishes a collaborative design process. It does not contain any executable scripts, network operations, or sensitive data access patterns.
[PROMPT_INJECTION]: The skill uses instructional tags like <HARD-GATE> to enforce design constraints rather than bypassing safety protocols. There are no attempts to override the underlying AI's safety guidelines or reveal system prompts.
[COMMAND_EXECUTION]: While the skill instructs the agent to 'explore project context' by checking files and commits, this is part of the standard operational behavior of a development assistant and does not involve arbitrary or malicious command execution.

brainstorm