The Agent Skills Directory

[COMMAND_EXECUTION]: The skill utilizes a Bash tool to perform Git operations. It includes explicit instructions to use quoted variables (e.g., "$branch") to protect against command injection from potentially malformed branch names.
[PROMPT_INJECTION]: No malicious patterns, bypass attempts, or role-play instructions were identified. The skill's logic is centered on a "safety first" approach that requires explicit user consent for all actions.
[DATA_EXFILTRATION]: The skill does not access sensitive files or perform non-standard network operations. Activities are restricted to local Git repository management and remote state synchronization via standard Git protocols.
[SAFE]: The skill follows security best practices by implementing two-stage verification gates through the AskUserQuestion tool and programmatically excluding protected branches from analysis.

git-cleanup