Write Perfect Conventional Commit Messages From Diffs
Paste your diff and get clear, conventional commit messages that tell the story of your changes.
Describe your git disaster in plain English, get the exact commands to fix it safely.
You are a git expert who has recovered from every possible git disaster. I messed something up and need help. What happened: [DESCRIBE THE MESS IN PLAIN ENGLISH] What I want: [WHAT THE FINAL STATE SHOULD BE] Am I on a team? [SOLO / TEAM — have others pulled?] Have I pushed? [YES / NO] Deliver: 1. **DIAGNOSIS** - What actually happened (in git terms) - Danger level: 🟢 Safe / 🟡 Careful / 🔴 Risky - Can this be undone? Fully / partially / data may be lost 2. **THE FIX** (exact commands) - Step-by-step commands with explanations - Safety checks to run before each step - Point of no return warning (if any) 3. **SAFETY NET** - Backup command to run FIRST (before any fix) - How to undo the fix if it makes things worse - How to verify the fix worked 4. **PREVENTION** - What to do differently next time - Git aliases or config to prevent this - Useful git commands to know IMPORTANT: Always start with `git status` and `git log --oneline -10` output. Ask me to run these if I haven't shared them. Never suggest `git push --force` without explicit warning about team impact.
Git disasters feel catastrophic but are almost always recoverable. This prompt provides exact commands for any git situation—from accidental force pushes to detached HEAD states—because it maps common panic scenarios to their specific solutions.
Use when you've accidentally committed to the wrong branch, when a rebase went wrong and you need to recover, when you've lost commits or need to undo a force push, or anytime git has put you in a state you don't know how to escape.
You'll get the exact git commands to run in order, with explanations of what each command does, safety checks before destructive operations, and verification steps to confirm the fix worked—taking you from panic to resolution in minutes.
Paste your diff and get clear, conventional commit messages that tell the story of your changes.
Don't just fix the bug. Understand why it happened, how to prevent it, and what else might break.
Generate a professional, blameless postmortem document from an incident. Learn from failures without finger-pointing.
Create a safe, step-by-step refactoring plan for messy legacy code without breaking production.
Turn vague 'it doesn't work' into a systematic debugging process that finds the root cause fast.