The Agent Skills Directory

[SAFE]: No malicious patterns or security risks were identified. The skill performs standard UX design tasks using local project files and user-provided descriptions.
[INDIRECT_PROMPT_INJECTION]: The skill ingests user input via arguments ($0 and $ARGUMENTS) and reads local project files (.styleseed/ project data, CLAUDE.md, and DESIGN-LANGUAGE.md) to generate design artifacts. While it lacks explicit boundary markers or sanitization for this input, the intended use case is creative design and the associated capabilities (Read, Write, Bash) are scoped to the project environment, representing a standard operational risk for this type of agent skill.

ss-flow