The Agent Skills Directory

[COMMAND_EXECUTION]: The skill instructs the agent to execute a shell command (ls) against various hidden local directories (e.g., ~/.claude/skills/, ~/.agent/skills/) to discover which other skills are currently installed. This is used for environment probing and automated cross-promotion of the author's other skills.
[EXTERNAL_DOWNLOADS]: The skill recommends that users install additional design-related skills from the author's repository (AbsolutelySkilled) using npx commands.
[PROMPT_INJECTION]: The skill possesses a vulnerability surface for indirect prompt injection as it is designed to process and style user-supplied UI code and components without using boundary markers or 'ignore' instructions.
Ingestion points: The skill ingests and processes user-provided UI components, CSS, and design tasks as described in SKILL.md and the reference files.
Boundary markers: No specific delimiters or safety instructions are present in the prompt to separate user-provided content from the agent's internal logic.
Capability inventory: The agent can execute filesystem discovery commands (ls) and make recommendations for external package installations (npx).
Sanitization: There is no evidence of sanitization, validation, or filtering of user-provided code before it is processed by the agent.

ultimate-ui