cleaning-up-branches
Pass
Audited by Gen Agent Trust Hub on Feb 17, 2026
Risk Level: SAFE
Full Analysis
- [COMMAND_EXECUTION] (SAFE): The skill utilizes standard Bash and Git commands to perform repository maintenance. All destructive operations (deleting local and remote branches) are gated behind explicit user confirmation prompts and support a dry-run flag for safe previews.
- [DATA_EXFILTRATION] (SAFE): Network activity is restricted to
git fetchandgit pushoperations directed at the user's own configured Git remotes (e.g., origin). No sensitive data or local files are transmitted to unauthorized third-party servers. - [REMOTE_CODE_EXECUTION] (SAFE): The skill does not download, install, or execute scripts from remote sources. All execution logic is local and uses established system tools.
- [INDIRECT_PROMPT_INJECTION] (SAFE): The skill processes untrusted data from the repository, such as branch names and commit messages. Security is maintained through consistent shell variable quoting (e.g.,
"$branch") which prevents command injection, and the requirement for manual user approval ensures that the AI agent does not act autonomously on potentially malicious repository metadata.
Audit Metadata