The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes existing theme and token files from the environment to generate its output, creating a surface where instructions hidden in those files could influence the agent.
Ingestion points: The execution flow requires reading existing tokens or theme files (SKILL.md).
Boundary markers: No delimiters or explicit instructions are provided to the agent to disregard instructions found within the ingested data.
Capability inventory: The skill uses a uiux_generate_tokens tool and produces configuration files in formats like Tailwind, CSS, and JSON.
Sanitization: There is no documentation of schema validation or content sanitization for the input data.

design-tokens