The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes data from references/cases.md which includes detailed prompt excerpts used as evidence for design patterns. This creates a surface where the agent might interpret these excerpts as active instructions.
Ingestion points: The references/cases.md file is accessed on every invocation to select a design anchor.
Boundary markers: The workflow requires a "Preserve / Replace / Avoid" contract to help the agent distinguish between reference data and user requirements.
Capability inventory: The skill can generate implementation-ready code, page narratives, and interaction specifications.
Sanitization: No specific content sanitization is performed on the ingested reference material.
[EXTERNAL_DOWNLOADS]: The skill provides links to external websites and assets for visual reference and documentation.
Evidence: Links to goodcase.ai, media.goodcase.ai, and x.com are included in the case evidence file for attribution and finished media inspection.

landing-page-information-structure-by-197q59i