schema-markup-generator

Pass

Audited by Gen Agent Trust Hub on Aug 31, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data from external web pages, including product descriptions, brand names, and existing schema markup.
  • Ingestion points: Data is collected from the page type, platform, canonical URL, product name, description, and existing theme/app markup (SKILL.md).
  • Boundary markers: The skill uses instructional markers such as "Require verified facts" and "Serialize strict JSON-LD" but does not specify clear delimiters (like XML tags or random hashes) for the external content in the logic.
  • Capability inventory: The skill is primarily a code generator; it does not contain subprocess calls or network write operations beyond the output provided to the user.
  • Sanitization: The workflow requires the agent to "Serialize strict JSON-LD and parse-check it" and "Compare every value with visible page content," which provides a layer of validation against injected instructions.
  • [EXTERNAL_DOWNLOADS]: The skill's installation process utilizes npx to fetch a package from the author's own repository (nexscope-ai/ecommerce-seo-geo-skills). This is a standard installation method for resources provided by the skill vendor.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 31, 2026, 05:39 AM
Security Audit — agent-trust-hub — schema-markup-generator