The Agent Skills Directory

[SAFE]: The skill is entirely instructional and informational, providing a multi-layer pipeline for reviewing Chinese text without any executable code or malicious scripts.
[SAFE]: All external references and checklists (advertising law, political compliance, etc.) are provided as static markdown files intended as context for the AI agent.
[COMMAND_EXECUTION]: The skill documentation maps interactive tools for user communication and file editing (e.g., Write, Edit, ask_user), which are appropriate for its stated purpose of reviewing and correcting text.
[SAFE]: No obfuscation, data exfiltration patterns, or unauthorized network operations were identified. The network capability listed in the metadata is associated with the stated fact-checking methodology.

editorial-review