Summary
- Tools: Husky + lint-staged enforce pre-commit standards.
- Mindset: Excuses for skipping tests (boring, time, tools) vs reality (discipline + setup).
- QA role: Combining unit testing, mocking, refactoring practices, and AI for coverage.
- Personal: Compare his preaching with my ATTEND method; check if mixing techniques yields stronger balance. Use as a mirror to test my ego.
Husky & Lint-Staged
- Automate pre-commit checks.
- Prevent “dirty” code from ever entering main branch.
- Pairs with formatting (Prettier) + linting (ESLint).
Excuses vs Reality
- “Too boring” → actually highlights weak spots.
- “No time” → cost of late bugs > time to test.
- “Doesn’t catch everything” → tests ≠ full coverage, but raise confidence baseline.
- “Hard to change later” → refactor + modularize to improve testability.
QA + AI in Testing
- QA moves from manual bug-finding to system validation + coverage guidance.
- AI helps:
- Suggest unit tests from code.
- Detect missing branches.
- Aid mocking and refactor impact analysis.
Personal Reflection
- Use his doctrine as a dilionário → compare what he preaches with my ATTEND approach (mixing techniques + pragmatism).
- Write down where they align and where I diverge.
- This helps control ego: I don’t discard others’ techniques just because I have my own system.
Practical Next Steps
- Implement Husky + lint-staged in at least one repo.
- Create a test excuse log: whenever I avoid a test, write why → then refactor it away.
- Try AI-assisted test generation for one module.
- Review ATTEND vs Steve’s approach → note pros/cons.