The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill instructs the agent to fetch repository metadata and file listings from the GitHub API. This involves accessing a well-known service to retrieve public information regarding the source and activity of external skills.
[COMMAND_EXECUTION]: Provides examples for using shell commands like curl and jq to retrieve and parse JSON data from the GitHub API. These commands are used for informational purposes and do not execute remote scripts.
[INDIRECT_PROMPT_INJECTION]: The skill's primary function is to ingest and evaluate third-party skill content. This creates an attack surface where untrusted instructions within the analyzed skills could attempt to subvert the vetting process or influence the agent's conclusions. The skill mitigates this by providing a mandatory code review checklist and risk classification rubric.

skill-vetter