What problem does it solve? AI coding agents can execute destructive git operations like force pushes, hard resets, and branch deletions that permanently destroy work. This Skill installs a PreToolUse hook that intercepts and blocks these commands before Claude Code runs them. ## Core Features & Use Cases - Command Interception: Blocks git push (including --force), git reset --hard, git clean -f/-fd, git branch -D, and git checkout/restore of the working tree via a bash hook script. - Flexible Scope: Installs the guardrail either per-project (.claude/settings.json) or globally (~/.claude/settings.json) for all projects. - Customizable Patterns: The blocked pattern list in the hook script can be edited to add or remove commands. - Use Case: A team lets Claude Code run autonomously on a repository but wants to guarantee it can never force-push to main or wipe uncommitted changes; the hook exits with code 2 and returns a BLOCKED message to the agent. ## Quick Start Ask the AI to set up git guardrails that block dangerous git commands like push and reset --hard in Claude Code for this project.
