ce-seo
Pass
Audited by Gen Agent Trust Hub on Aug 22, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill describes functionality for generating Markdown mirrors and 'llms.txt' files from product catalog data (names, descriptions, attributes) to be consumed by AI agents.
- Ingestion points: External data enters the system from the Commerce Engine storefront API (via the
ce-catalogreference). - Boundary markers: The documentation explicitly states that Markdown escaping is applied to special characters such as brackets and pipes to prevent structural or markup injection in the generated mirrors.
- Capability inventory: The skill provides instructions for tools to generate and serve text-based SEO assets; it does not execute arbitrary code on the content itself.
- Sanitization: Serialization of JSON-LD and Markdown escaping are used to ensure data integrity.
- [EXTERNAL_DOWNLOADS]: The skill references the installation of the
@commercengine/seopackage and fetching markdown documentation fromllm-docs.commercengine.io. These are official resources owned by the skill's author (commercengine).
Audit Metadata