The Agent Skills Directory

[SAFE]: No malicious patterns, obfuscation, or suspicious code execution sequences were found. The skill operates according to its stated purpose of content generation.
[SAFE]: The skill performs authorized file system operations including reading user-defined style guides from the '10 About me' directory and saving deliverables to the '08 交付物' directory.
[PROMPT_INJECTION]: The skill processes untrusted user data (briefs), creating a surface for indirect prompt injection. 1. Ingestion points: User-provided commercial briefs or product descriptions entering via the interaction prompt. 2. Boundary markers: None are specified to isolate the brief content from the system instructions. 3. Capability inventory: The skill is capable of writing content to the local file system. 4. Sanitization: The skill explicitly instructs the agent to convert confidential information from the brief into vague expressions and avoid exposing internal data, which serves as a task-specific sanitization measure.

branded-content