# Small Steps and Iteration — AI Coding Assistants: Use Them Well and Review Them Hard

Source: https://www.geekswithgeeks.com/en/ai-coding-assistants/prompt-iterate

> Break work into small, reviewable changes and correct the assistant when it drifts.

## Small diffs are reviewable

Asking for a whole feature at once often returns a large change that is hard to review and easy to get subtly wrong. Instead ask for **one step at a time**: the data model, then the function, then the endpoint, then the tests. Commit after each working step so you can undo easily. If the assistant goes in a wrong direction, say what is wrong and what to keep rather than starting over from scratch, and restart with a fresh chat if the conversation gets long and confused.

## A step-by-step plan

Each step ends with something you can run and check.

```text
Step 1  Add the Reminder model + migration           -> run migration
Step 2  Write tests for create / list / delete            -> tests fail (no code yet)
Step 3  Implement the service functions                   -> tests pass
Step 4  Add the REST endpoints                            -> curl each one
Step 5  Review the full diff, then commit
```

**Quiz:** Why ask for a feature in small steps?

- [ ] It is required by Git
- [ ] Assistants cannot write large code
- [x] Small changes are easier to review, test and undo
- [ ] It saves no time ever

*Answer:* Small changes are easier to review, test and undo. Large, unreviewed diffs hide mistakes; small verified steps keep you in control.
