The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill provides commands to fetch repository metadata and file contents from GitHub and GitHub User Content. These operations target well-known services and are used for the legitimate purpose of retrieving code for security inspection.
[INDIRECT_PROMPT_INJECTION]: As a security tool, this skill is designed to ingest and analyze untrusted skill code from external sources. This creates a surface where the agent might encounter malicious instructions embedded in the reviewed content.
Ingestion points: SKILL.md (instructions to read external skill files and fetch content via curl).
Boundary markers: The protocol does not define specific delimiters for the external content being analyzed.
Capability inventory: curl for network retrieval and jq for parsing metadata. No automated execution of the retrieved code is performed.
Sanitization: The skill's primary function is the detection and manual/reasoned analysis of potentially malicious content before any installation or execution occurs.

skill-vetter