The Agent Skills Directory

[SAFE]: The skill's primary logic is focused on content auditing and text transformation based on a provided local reference file. There are no instructions or scripts that involve network communication, credential harvesting, or system privilege modification.
[SAFE]: The external links included in the knowledge base are informational in nature, pointing to official platform guidelines, regulatory news articles, and established marketing utility sites. No automated scripts or tools are configured to fetch or execute data from these remote sources.
[PROMPT_INJECTION]: The skill identifies a potential surface for indirect prompt injection because it is designed to ingest and process untrusted user-supplied content (social media copy) as its main function.
Ingestion points: User-provided notes and文案 (copy) during the interactive checking process (SKILL.md).
Boundary markers: The instructions do not define specific delimiters or include guidance to the agent to ignore instructions embedded within the user data being audited.
Capability inventory: The skill does not possess or utilize any dangerous tools, such as network fetching, file writing, or subprocess execution capabilities across any of its provided files.
Sanitization: There is no evidence of input sanitization or filtering to prevent the agent from potentially responding to instructions contained within the user-provided text.

jacky-xhs-check