The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill incorporates user-supplied data into its generation logic, creating a surface for potential indirect prompt injection.
Ingestion points: The $0 (pattern type) and $ARGUMENTS (description) variables in SKILL.md are used to capture user input.
Boundary markers: No specific boundary markers or instructions to treat input as non-executable data are provided.
Capability inventory: The skill is designed to read local files (e.g., CLAUDE.md, components/ui/, and components/patterns/) and generate code based on those contents.
Sanitization: There is no evidence of input validation or sanitization for the data passed through the placeholders.
[METADATA_POISONING]: The YAML frontmatter includes a risk: critical field. This metadata is inconsistent with the skill's benign UI-generation purpose and could be misleading to users or automated security systems.

ui-pattern