The Agent Skills Directory

[COMMAND_EXECUTION]: The skill utilizes the npx ai-agent-skills tool to perform administrative actions such as previewing content, validating structure, and curating skill metadata within a catalog.
[INDIRECT_PROMPT_INJECTION]: The skill identifies a workflow for processing external skill data, creating a potential surface for indirect prompt injection. * Ingestion points: External skills are ingested via the preview and validate commands in SKILL.md. * Boundary markers: The instructions explicitly advise the agent to monitor for sanitization flags and qualitatively evaluate safety. * Capability inventory: The skill uses CLI tools to manage the lifecycle of other agent skills. * Sanitization: Relies on the automated sanitization provided by the preview command and subsequent agent/human validation.

review-a-skill