The Testing Loop

AI code needs verification. Build this habit.

The Verification Flow

AI generates code

Get the initial implementation.

You read it (actually read it)

Don't just glance. Understand what it does.

Run it locally

npm run dev

Check edge cases

What happens with empty input? Invalid data? Network errors?

Run tests

npm run test

Build passes

npm run build

Only then commit

git add .
git commit -m "Add feature X"

Test-Driven Vibecoding

Even better — write tests first:

Step 1: "Write a test for a function that validates email addresses"
        → AI writes test

Step 2: "Now implement the function to make the test pass"
        → AI writes implementation

Step 3: Run test → verify it passes

Step 4: "Add edge case tests for: empty string, missing @, multiple @s"
        → AI adds more tests

Step 5: Iterate until solid

Tests give AI a clear target and catch regressions.

Quick Sanity Checks

Before accepting AI code, verify:

Does this actually solve my problem?
Are there any obvious security issues?
Does it handle errors?
Is it consistent with the rest of the codebase?
Would I be embarrassed to show this to a senior dev?

Build Before Commit (Always)

# Make this a habit - every single time
npm run build    # or your build command
npm run test     # run test suite
npm run lint     # check for issues
 
# Only if all pass:
git add .
git commit -m "Add feature X"

🚫

Never commit broken code. If build fails, fix it before committing.

Common Test Prompts

"Write unit tests for the validateEmail function covering:
- Valid emails
- Empty string
- Missing @ symbol
- Missing domain
- Multiple @ symbols"

"Create integration tests for the /api/users endpoint:
- GET returns list of users
- POST creates new user
- POST with invalid data returns 400
- Unauthorized request returns 401"

"Add test coverage for error cases in the PaymentService"

Common Mistakes Security Mindset