The Agent Skills Directory

[SAFE]: The skill defines a transparent and legitimate workflow for assisting users with product reviews. It implements security best practices by mandating a human-in-the-loop approval step before any state-changing actions are performed.
[INDIRECT_PROMPT_INJECTION]: The skill interacts with external web content from Amazon product pages and order histories, which are potential sources for untrusted data. However, this surface is mitigated by the requirement that users must verify and approve the final review text before submission.

write-reviews