docs/PRE_COMMIT_CHECKLIST.md

# Pre-Commit Checklist

**MANDATORY**: Check ALL items before every commit.

## 1. Task Validation
- [ ] Task is listed in `planning.md`
- [ ] All task dependencies are complete
- [ ] Prerequisites are verified
- [ ] Success criteria understood

## 2. Code Reading (Before Implementation)
- [ ] Used `grep_search` to find related code
- [ ] Read existing function signatures with `read_file`
- [ ] Verified function return types
- [ ] Checked parameter names and types
- [ ] No assumptions made about signatures

## 3. Code Quality
- [ ] **Run linter**: `conda run -n ai_engg ruff check . --fix`
- [ ] **Fix ALL errors** (especially):
  - Line length (88 character limit)
  - Import order (stdlib → third-party → local)
  - Unused imports
  - Formatting issues
- [ ] Code tested and works
- [ ] Type hints added to new functions
- [ ] Docstrings added to new functions

## 4. Documentation
- [ ] Update `planning.md` if task complete (mark ✅)
- [ ] Commit message follows format: `type: brief summary`
- [ ] Commit is atomic (one logical change)

## 5. Git
- [ ] `git status` checked
- [ ] Only relevant files staged
- [ ] No debug files or `.pyc` included
- [ ] No secrets or credentials

## Common Commit Types
- `feat:` - New feature
- `fix:` - Bug fix
- `refactor:` - Code restructuring
- `docs:` - Documentation only
- `test:` - Adding/updating tests
- `chore:` - Maintenance tasks

## Recent Mistakes to Avoid
- ❌ Skipping task dependency check (did 2.3 before 2.2b)
- ❌ Assuming function signatures (3 errors in train_with_mlflow.py)
- ❌ Not running ruff (71 errors accumulated)
- ❌ Committing without testing

## Quick Commands
```bash
# Check code quality
conda run -n ai_engg ruff check . --fix

# Run specific test
conda run -n ai_engg python scripts/test_module.py

# Check git status
git status

# Stage and commit
git add <files>
git commit -m "type: message"
```

---
**Remember**: 30 seconds of reading prevents 5 minutes of debugging!