The Agent Skills Directory

[SAFE]: The skill performs standard administrative tasks focused on brand identity management. It manages content within local workspace directories such as .agents/ or .claude/ and does not demonstrate any malicious intent or unauthorized access patterns.
[DATA_EXFILTRATION]: No network activity was identified. The skill captures user input and stores it locally; it does not contain any curl, wget, or other network-based tools that would suggest data exfiltration.
[REMOTE_CODE_EXECUTION]: The skill does not download or execute external scripts, binary files, or packages. It relies purely on markdown-based instructions for data collection.
[PROMPT_INJECTION]: The instructions are clear and descriptive, focusing on guiding the agent through a structured interview process with the user without attempting to bypass safety filters or override system constraints.

brand-context