The Agent Skills Directory

[COMMAND_EXECUTION]: The skill provides instructions for a wide range of Git commands, including destructive operations. It mitigates potential risks by enforcing safety protocols such as checking status, creating backup branches, and requiring explicit user confirmation before any destructive action is performed.
[EXTERNAL_DOWNLOADS]: The documentation references the installation of well-known, official development tools including the GitHub CLI (gh), Azure DevOps CLI (az), GitLab CLI (glab), and the git-filter-repo utility. These are standard, reputable tools within the version control ecosystem.
[INDIRECT_PROMPT_INJECTION]: The skill processes repository data and file contents, creating an attack surface for indirect prompt injection. This risk is addressed through the following evidence chain: (1) Ingestion points: The skill interacts with local files, repository history, and remote branch data. (2) Boundary markers: Instructions explicitly mandate asking for user preference at the start of any task and include confirmation prompts for risky actions. (3) Capability inventory: The skill can execute shell commands via Git, modify repository configuration, and manage Git hooks. (4) Sanitization: Safety protocols include manual review steps and confirmation scripts for high-risk operations to ensure user oversight.
[SAFE]: The skill includes dedicated guidance on security best practices, such as proper credential management, SSH/GPG configuration, and the implementation of hooks to detect and prevent the accidental commitment of secrets.

git-master