The Agent Skills Directory

[SAFE]: The skill enforces a mandatory user-confirmation step for all sensitive operations, including git commits, merges, and branch deletions, preventing unauthorized or accidental actions.
[SAFE]: Explicitly instructs the agent to check for and prevent the commitment of secret-bearing files such as .env or credential files.
[SAFE]: Processes external context from git diffs, Jira tickets, and local specifications to generate summaries, but uses a 'human-in-the-loop' design to ensure that all generated text is reviewed by the user before being applied or shared.

deploy-ship