The Agent Skills Directory

[SAFE]: The skill implements strong security guidelines for handling external content. It explicitly instructs the agent to treat all fetched URLs as untrusted input and specifically forbids following instructions or executing scripts embedded in page content.
[SAFE]: Network operations are restricted and well-defined. The agent is instructed to fetch only user-provided URLs and specific variations required for analysis, with explicit rules against following redirects to unauthorized domains.
[SAFE]: The skill avoids high-risk operations. There are no instructions for local file system writes, credential harvesting, or the creation of persistence mechanisms. The primary function is text analysis and content generation within the authoring constraints of Adobe Edge Delivery Services.

geo-rewrite