This skill provides a structured 'finish mode' for AI agents. It replaces broad git add . commands with a cautious workflow: inspecting status, running local verification tests, explicitly staging intended files, and pushing changes.
Use this skill when a task is complete and the agent is asked to 'commit and push', '/cp', or 'upload changes to main'. It is designed to prevent accidental pushes of secrets, unrelated local changes, or unverified code.
Primarily designed for Codex and other CLI-based AI agents with shell access.
This skill has not been reviewed by our automated audit pipeline yet.