test-driven-development

Test-Driven Development (TDD)

Overview

Write the test first. Watch it fail. Write minimal code to pass.

Core principle: If you didn't watch the test fail, you don't know if it tests the right thing.

Violating the letter of the rules is violating the spirit of the rules.

When to Use

Always:

• New features
• Bug fixes
• Refactoring
• Behavior changes

Exceptions (ask the user):

• Throwaway prototypes
• Generated code
• Configuration files

Thinking "skip TDD just this once"? Stop. That's rationalization.

The Iron Law

NO PRODUCTION CODE WITHOUT A FAILING TEST FIRST

Write code before the test? Delete it. Start over.

No exceptions:

• Don't keep it as "reference"
• Don't "adapt" it while writing tests
• Don't look at it
• Delete means delete

Implement fresh from tests. Period.

Red-Green-Refactor

digraph tdd_cycle {
    rankdir=LR;
    red [label="RED\nWrite failing test", shape=box, style=filled, fillcolor="#ffcccc"];
    verify_red [label="Verify fails\ncorrectly", shape=diamond];
    green [label="GREEN\nMinimal code", shape=box, style=filled, fillcolor="#ccffcc"];
    verify_green [label="Verify passes\nAll green", shape=diamond];
    refactor [label="REFACTOR\nClean up", shape=box, style=filled, fillcolor="#ccccff"];
    next [label="Next", shape=ellipse];

    red -> verify_red;
    verify_red -> green [label="yes"];
    verify_red -> red [label="wrong\nfailure"];
    green -> verify_green;
    verify_green -> refactor [label="yes"];
    verify_green -> green [label="no"];
    refactor -> verify_green [label="stay\ngreen"];
    verify_green -> next;
    next -> red;
}

RED - Write Failing Test

Write one minimal test showing what should happen.

```typescript test('retries failed operations 3 times', async () => { let attempts = 0; const operation = () => { attempts++; if (attempts < 3) throw new Error('fail'); return 'success'; };

const result = await retryOperation(operation);

expect(result).toBe('success'); expect(attempts).toBe(3); });

Clear name, tests real behavior, one thing
</Good>

<Bad>
```typescript
test('retry works', async () => {
  const mock = jest.fn()
    .mockRejectedValueOnce(new Error())
    .mockRejectedValueOnce(new Error())
    .mockResolvedValueOnce('success');
  await retryOperation(mock);
  expect(mock).toHaveBeenCalledTimes(3);
});

Vague name, tests mock not code

Requirements:

• One behavior
• Clear name
• Real code (no mocks unless unavoidable)

Verify RED - Watch It Fail

MANDATORY. Never skip.

npm test path/to/test.test.ts

Confirm:

• Test fails (not errors)
• Failure message is expected
• Fails because feature missing (not typos)

Test passes? You're testing existing behavior. Fix test.

Test errors? Fix error, re-run until it fails correctly.

GREEN - Minimal Code

Write simplest code to pass the test.

```typescript async function retryOperation(fn: () => Promise): Promise { for (let i = 0; i < 3; i++) { try { return await fn(); } catch (e) { if (i === 2) throw e; } } throw new Error('unreachable'); } ``` Just enough to pass ```typescript async function retryOperation( fn: () => Promise, options?: { maxRetries?: number; backoff?: 'linear' | 'exponential'; onRetry?: (attempt: number) => void; } ): Promise { // YAGNI } ``` Over-engineered

Don't add features, refactor other code, or "improve" beyond the test.

Verify GREEN - Watch It Pass

MANDATORY.

npm test path/to/test.test.ts

Confirm:

• Test passes
• Other tests still pass
• Output pristine (no errors, warnings)

Test fails? Fix code, not test.

Other tests fail? Fix now.

REFACTOR - Clean Up

After green only:

• Remove duplication
• Improve names
• Extract helpers

Keep tests green. Don't add behavior.

Repeat

Next failing test for next feature.

Instrumented TDD — Logs as Breakpoints

Every test you write and every line of production code you implement should be instrumented with structured JSONL logging. Trace logs are your breakpoints — they dump variable values so AI can see exactly what happened when a test fails.

Prerequisite: If the codebase doesn't have structured logging yet, run debugging:logging-enablement first to set up the logger, JSONL file sink, and test harness integration.

Log Levels in TDD Code

Level	Where	What	EUII	Production
Trace	Production code	Variable values, intermediate state, input/output of functions — the stuff you'd put a breakpoint on	YES (stripped from release builds)	Compiled out, zero overhead
Debug	Production code	Positive handshakes: "OrderService.Place entered", "validation passed". Narrows WHERE, not WHAT	NO	Kept, toggled on demand
Information	Production code	Business event sequence: "Order accepted", "Payment processed". Execution flow for timeline reconstruction	NO	Always on
Warning	Production code	Unexpected but recoverable: retry, fallback, degraded mode	NO	Always on
Error	Production code	Operation failure, recoverable at higher level	NO	Always on

Tests themselves log at Debug/Info to mark test phases (arrange/act/assert) and capture outcomes.

Why Trace ≠ Production Risk

Trace is compiled out of release builds. This is the only reason EUII is safe at Trace:

• Local dev / tests: Trace minimum → full variable visibility
• Production: Information minimum → no EUII, no noise
• Release builds: Trace statements produce zero overhead

RED Phase — Instrumented Tests

When writing the failing test, set up a structured logger with test context:

test('retries failed operations 3 times', async () => {
  // Bind logger with test-case-name inside each test
  const log = withTestCase(moduleLogger, 'retries failed operations 3 times');

  let attempts = 0;
  const operation = () => {
    attempts++;
    log.trace({ attempt: attempts }, 'Operation attempt');
    if (attempts < 3) throw new Error('fail');
    return 'success';
  };

  const result = await retryOperation(operation, log);

  log.debug({ result, attempts }, 'Assert: verifying retry behavior');
  expect(result).toBe('success');
  expect(attempts).toBe(3);
});

GREEN Phase — Instrumented Production Code

When writing minimal code to pass the test, add logging at decision points:

async function retryOperation<T>(fn: () => Promise<T>, log: Logger): Promise<T> {
  for (let i = 0; i < 3; i++) {
    try {
      log.trace({ attempt: i + 1 }, 'Attempting operation');
      const result = await fn();
      log.debug({ attempt: i + 1 }, 'Operation succeeded');
      return result;
    } catch (e) {
      log.trace({ attempt: i + 1, error: e.message }, 'Operation failed, will retry');
      if (i === 2) {
        log.error({ attempt: i + 1, error: e.message }, 'All retries exhausted');
        throw e;
      }
    }
  }
  throw new Error('unreachable');
}

Key rules:

• Trace for variable dumps (attempt number, intermediate values, caught errors) — breakpoint equivalents
• Debug for positive handshakes ("operation succeeded") — narrows WHERE
• Info for business events ("order accepted") — timeline reconstruction
• Error for failures the caller will handle
• EUII (emails, user names, IPs) is Trace-only — forbidden at Debug and above

When a Test Fails Unexpectedly

Don't guess. Read the logs.

1. Run the failing test
2. Open the JSONL log file (e.g., my-app.log.jsonl)
3. Query with DuckDB to see exactly what happened:

-- What happened during the failing test?
SELECT "@t", "@l", "@m"
FROM read_json_auto('my-app.log.jsonl')
WHERE "test-case-name" = 'retries failed operations 3 times'
ORDER BY "@t";

-- What were the variable values at Trace level?
SELECT "@t", "@m", attempt, error
FROM read_json_auto('my-app.log.jsonl')
WHERE "test-case-name" = 'retries failed operations 3 times'
  AND "@l" = 'Trace'
ORDER BY "@t";

4. If logs don't reveal the issue → add more Trace logging at the gap → re-run → re-query
5. For deeper investigation, use debugging:debug-with-logs skill

The logs replace the debugger. Trace-level logs dump every variable you'd inspect at a breakpoint. The AI reads the JSONL and sees the full execution trace.

Good Tests

Quality	Good	Bad
Minimal	One thing. "and" in name? Split it.	test('validates email and domain and whitespace')
Clear	Name describes behavior	test('test1')
Shows intent	Demonstrates desired API	Obscures what code should do

Why Order Matters

"I'll write tests after to verify it works"

Tests written after code pass immediately. Passing immediately proves nothing:

• Might test wrong thing
• Might test implementation, not behavior
• Might miss edge cases you forgot
• You never saw it catch the bug

Test-first forces you to see the test fail, proving it actually tests something.

"I already manually tested all the edge cases"

Manual testing is ad-hoc. You think you tested everything but:

• No record of what you tested
• Can't re-run when code changes
• Easy to forget cases under pressure
• "It worked when I tried it" ≠ comprehensive

Automated tests are systematic. They run the same way every time.

"Deleting X hours of work is wasteful"

Sunk cost fallacy. The time is already gone. Your choice now:

• Delete and rewrite with TDD (X more hours, high confidence)
• Keep it and add tests after (30 min, low confidence, likely bugs)

The "waste" is keeping code you can't trust. Working code without real tests is technical debt.

"TDD is dogmatic, being pragmatic means adapting"

TDD IS pragmatic:

• Finds bugs before commit (faster than debugging after)
• Prevents regressions (tests catch breaks immediately)
• Documents behavior (tests show how to use code)
• Enables refactoring (change freely, tests catch breaks)

"Pragmatic" shortcuts = debugging in production = slower.

"Tests after achieve the same goals - it's spirit not ritual"

No. Tests-after answer "What does this do?" Tests-first answer "What should this do?"

Tests-after are biased by your implementation. You test what you built, not what's required. You verify remembered edge cases, not discovered ones.

Tests-first force edge case discovery before implementing. Tests-after verify you remembered everything (you didn't).

30 minutes of tests after ≠ TDD. You get coverage, lose proof tests work.

Common Rationalizations

Excuse	Reality
"Too simple to test"	Simple code breaks. Test takes 30 seconds.
"I'll test after"	Tests passing immediately prove nothing.
"Tests after achieve same goals"	Tests-after = "what does this do?" Tests-first = "what should this do?"
"Already manually tested"	Ad-hoc ≠ systematic. No record, can't re-run.
"Deleting X hours is wasteful"	Sunk cost fallacy. Keeping unverified code is technical debt.
"Keep as reference, write tests first"	You'll adapt it. That's testing after. Delete means delete.
"Need to explore first"	Fine. Throw away exploration, start with TDD.
"Test hard = design unclear"	Listen to test. Hard to test = hard to use.
"TDD will slow me down"	TDD faster than debugging. Pragmatic = test-first.
"Manual test faster"	Manual doesn't prove edge cases. You'll re-test every change.
"Existing code has no tests"	You're improving it. Add tests for existing code.

Red Flags - STOP and Start Over

• Code before test
• Test after implementation
• Test passes immediately
• Can't explain why test failed
• Tests added "later"
• Rationalizing "just this once"
• "I already manually tested it"
• "Tests after achieve the same purpose"
• "It's about spirit not ritual"
• "Keep as reference" or "adapt existing code"
• "Already spent X hours, deleting is wasteful"
• "TDD is dogmatic, I'm being pragmatic"
• "This is different because..."

All of these mean: Delete code. Start over with TDD.

Example: Bug Fix

Bug: Empty email accepted

RED

test('rejects empty email', async () => {
  const result = await submitForm({ email: '' });
  expect(result.error).toBe('Email required');
});

Verify RED

$ npm test
FAIL: expected 'Email required', got undefined

GREEN

function submitForm(data: FormData) {
  if (!data.email?.trim()) {
    return { error: 'Email required' };
  }
  // ...
}

Verify GREEN

$ npm test
PASS

REFACTOR Extract validation for multiple fields if needed.

Verification Checklist

Before marking work complete:

• Every new function/method has a test
• Watched each test fail before implementing
• Each test failed for expected reason (feature missing, not typo)
• Wrote minimal code to pass each test
• All tests pass
• Output pristine (no errors, warnings)
• Tests use real code (mocks only if unavoidable)
• Edge cases and errors covered
• Production code has Trace logging at decision points (variable values, branch entries)
• Production code has Debug/Info logging for execution flow (no EUII at these levels)
• Tests use structured logger with test-case-name and test-module-name context
• JSONL log output is queryable and shows full execution trace for each test

Can't check all boxes? You skipped TDD. Start over.

When Stuck

Problem	Solution
Don't know how to test	Write wished-for API. Write assertion first. Ask the user.
Test too complicated	Design too complicated. Simplify interface.
Must mock everything	Code too coupled. Use dependency injection.
Test setup huge	Extract helpers. Still complex? Simplify design.

Debugging Integration

Bug found? Logs first, then test.

1. Ensure logging is enabled — If the codebase lacks structured JSONL logging, run debugging:logging-enablement before anything else
2. Reproduce with Trace logging on — Run the buggy code path with minimum level set to Trace
3. Query the JSONL logs — Use DuckDB to trace the execution and find where values go wrong (use debugging:debug-with-logs for the full methodology)
4. Write failing test reproducing the bug — Now that you understand the root cause from the logs
5. Fix the code — Minimal change to pass the test
6. Verify via logs — Re-run and confirm the JSONL logs show correct behavior at the fix point

Never fix bugs without a test. Never diagnose bugs without reading the logs.

For complex or multi-attempt failures, use debugging:systematic-debugging — it enforces root cause investigation before any fix attempts.

Testing Anti-Patterns

When adding mocks or test utilities, read testing-anti-patterns.md to avoid common pitfalls:

• Testing mock behavior instead of real behavior
• Adding test-only methods to production classes
• Mocking without understanding dependencies

Final Rule

Production code → test exists and failed first
Otherwise → not TDD

No exceptions without the user's permission.