The Agent Skills Directory

[SAFE]: The skill is purely instructional and contains no code or commands. The primary file, SKILL.md, provides a template for the agent to use when asked to 'zoom out' and describe code at a higher level of abstraction.
[SAFE]: The metadata field disable-model-invocation: true explicitly prevents the skill from triggering model-based tool calls, further limiting its security profile.
[SAFE]: No obfuscation, data exfiltration, or prompt injection patterns were detected. The attribution in SOURCE.md points to a public repository for credit and versioning info, which is a standard practice.

zoom-out