The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill defines a surface for indirect prompt injection by instructing the agent to ingest and process data from external sources, such as design tokens and Figma metadata.
Ingestion points: Instructions in SKILL.md mandate using get_design_context and pulling metadata/screenshots from user-provided Figma URLs.
Boundary markers: The instructions do not specify any delimiters or warnings to ignore embedded instructions within the design tokens or Figma metadata.
Capability inventory: The skill provides HTML/CSS templates and layout rules but does not contain scripts with subprocess calls, file-write capabilities, or network operations.
Sanitization: There are no mentioned sanitization, validation, or filtering steps for the data retrieved from external design sources.

dashboard-design