The Agent Skills Directory

[SAFE]: The skill uses dynamic context injection (via the !command syntax) to retrieve repository metadata such as the current tag, commit history, and branch name. These are read-only git operations used appropriately for context gathering and do not incorporate untrusted user input into the shell commands.
[SAFE]: The skill processes repository commit messages to determine semantic versioning. It utilizes XML-style boundary markers (<commits_since_tag>) to isolate this external content from the instruction set. Given the skill's specific focus on metadata analysis and its limited capability set (git tag and push), the potential for indirect prompt injection is negligible and consistent with the tool's intended purpose.

tag