design-page
Pass
Audited by Gen Agent Trust Hub on Oct 12, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes potentially untrusted content from multiple project files (e.g., brand.md, products.md, and page plans) to author page source and platform configurations.
- Ingestion points: work/storefront/setup.md, brand.md, design.md, products.md, persona.md, rules.md, and various page plans.
- Boundary markers: Instructions specify placing plan copy verbatim and using exact Lexsis actions, but they do not define explicit prompt delimiters or "ignore embedded instructions" directives for the ingested markdown content.
- Capability inventory: The skill utilizes powerful tools including lexsis_page_create.create, lexsis_drafts.page_patch, and lexsis_pages.compile to create and modify drafts on a remote platform.
- Sanitization: The skill relies on downstream "schema-valid islands" and compiler errors for validation but lacks explicit sanitization or escaping of the input markdown text before interpolation into design source.
Audit Metadata