The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: Fetches repository information and skill files from GitHub's official API and raw content domains to facilitate security reviews.
[COMMAND_EXECUTION]: Provides bash commands using curl and jq to retrieve and parse structured data from external repositories for auditing.
[PROMPT_INJECTION]: Ingests untrusted markdown and JSON content from third-party repositories. While this presents an indirect prompt injection surface, the skill is specifically designed to facilitate manual review of such content to identify malicious patterns.

skill-vetter