The Agent Skills Directory

[SAFE]: The skill demonstrates high security standards for repository interaction. Technical controls include explicit prohibitions against forceful tag updates and deletions, as well as a requirement for clean working trees before creation. All consequential actions, particularly remote pushes (R4 risk level), are gated by explicit user confirmation. The use of lazy-loaded references ensures that specific Git command details are only introduced when relevant to the current task context. Analysis across all threat categories, including obfuscation and data exfiltration, confirms the absence of malicious intent or risky implementation details.

tag-release