The Agent Skills Directory

[SAFE]: The skill defines a structured workflow for project planning and design validation. It focuses on gathering requirements and documenting designs before implementation. No malicious patterns were detected.- [PROMPT_INJECTION]: The instructions use strong imperative language ("MUST", "Do NOT") to enforce a specific operational logic (a "Hard Gate"). This is a functional constraint to ensure user approval before code generation, not a malicious attempt to bypass model safety filters.- [DATA_EXPOSURE]: The skill instructs the agent to "Explore project context" by checking local files and git commits. This is standard behavior for a development assistant and lacks any network exfiltration mechanisms or access to sensitive system paths (e.g., ~/.ssh).- [COMMAND_EXECUTION]: The skill allows writing design documents to the docs/plans/ directory and committing them to git. These are restricted file-system operations consistent with the skill's stated purpose of documentation.

brainstorming