The Agent Skills Directory

[SAFE]: The skill provides instructional content for SEO and structured data implementation. It correctly identifies and mitigates a common security risk (XSS/Script injection) associated with embedding dynamic content into JSON-LD script blocks by recommending the replacement of < with the Unicode escape \u003c.
[SAFE]: The skill references standard, well-known industry tools including Google's Rich Results Test and the official Schema.org Validator. All external documentation links point to official domains such as google.com and nextjs.org.
[INDIRECT_PROMPT_INJECTION]: The skill is designed to handle untrusted data from a CMS (e.g., product names and descriptions) to generate structured data. While this presents a surface for indirect injection, the skill specifically instructs the user to sanitize this data to prevent breaking out of the HTML script boundary, which is a significant security-positive feature.
Ingestion points: Product names, descriptions, and FAQ content retrieved from a CMS or external source as described in SKILL.md and evals.json (Prompt 7).
Boundary markers: JSON-LD is wrapped in <script type="application/ld+json"> tags.
Capability inventory: The skill generates text and code snippets for implementation; it does not execute system commands or perform direct file writes outside of standard agent interaction.
Sanitization: Explicitly recommends escaping < characters as \u003c during JSON serialization to prevent malicious content from terminating the script block and executing arbitrary JavaScript.

schema