The Agent Skills Directory

Indirect Prompt Injection Surface: The skill is designed to audit and document UI components by ingesting data from external design tools and knowledge bases. This creates a surface where instructions embedded in design data could potentially influence agent behavior.
Ingestion points: Data retrieved from design tool connectors and knowledge bases, as well as user-supplied component definitions in SKILL.md arguments.
Boundary markers: The provided output templates do not currently include explicit markers or instructions to isolate external data from the agent's core processing logic.
Capability inventory: No executable scripts, shell commands, or network operations are defined within the skill's own instructions.
Sanitization: There are no explicit validation or sanitization routines described for the ingested design data.
No Executable Code: The skill consists entirely of Markdown instructions and templates without any associated scripts or binary files, which inherently limits the scope of potential execution risks.

design-system