test-driven-development

name	test-driven-development
description	Use when implementing any assigned scenario or bugfix, before writing implementation code

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.

shopsystem note (experimental). This is an adapted variant of the external test-driven-development skill, tuned for a BC Implementer running autonomously inside a container. There is no human in the BC's loop; where the original skill defers to a "human partner," the BC instead emits a clarify message to the lead and awaits the lead's decision. This draft is meant to be learned from, not treated as canon.

TDD Is Mandatory in This BC

TDD is MANDATORY in this BC — not optional. There is no self-granted exception. If you believe an exception applies (throwaway prototype, generated code, configuration file), you do NOT decide unilaterally: emit a clarify message to the lead naming the work and the exception you think applies, and await the lead's decision before proceeding without TDD.

The only exception path is: emit clarify → await the lead's decision. No other path exists.

Staged-Commit Convention

For each behavior, commit the cycle phases separately and with the prescribed prefixes. Frequent, staged commits are required — at minimum one RED commit and one GREEN commit per behavior:

Phase	Commit message	When to commit
RED	test(red): <behavior>	After writing the failing test, before any implementation
GREEN	feat(green): <behavior>	After the test passes with minimal implementation
REFACTOR	refactor: <behavior>	After each clean-up pass (optional but encouraged)

Never combine RED and GREEN into one commit. The commit history is an observable artifact — the router's inter-layer gate and the work-done-gate both verify that test(red): <behavior> precedes feat(green): <behavior> in the work-branch history.

Example (for behavior "empty email rejection"):

# After writing the failing test:
git add tests/test_email.py
git commit -m "test(red): empty email rejection"

# After implementing the fix:
git add src/email.py
git commit -m "feat(green): empty email rejection"

# Optional refactor:
git add src/email.py
git commit -m "refactor: empty email rejection"

The Two Loops

A BC Implementer works against two nested loops, and TDD lives in the inner one.

• Outer loop — the assigned Gherkin scenario(s). The scenario(s) the lead dispatched are the acceptance specification: they pin what behavior is required and they are what work_done proves. The scenario is silent on internal decomposition — it says nothing about which functions, classes, or modules you write. The outer loop is satisfied when the assigned scenario(s) pass against your implementation.

• Inner loop — RED-GREEN-REFACTOR. For each behavior you must build to make the assigned scenario(s) pass, you run the TDD cycle below. This is where the BC does its own internal design: the scenario tells you the destination, and the inner loop is how you get there, one failing test at a time. This skill fills the gap the scenario deliberately leaves open.

When a piece of assigned work decomposes into multiple behaviors, track those behaviors as bd sub-issues of the work's lead bead — never as TodoWrite entries or markdown checklists. The bd registry is the single source of truth for decomposition.

When to Use

Always:

• New features
• Bug fixes
• Refactoring
• Behavior changes

Exceptions (emit a clarify to the lead and await the decision):

• Throwaway prototypes
• Generated code
• Configuration files

When you believe an exception applies, do not decide unilaterally: emit a clarify message to the lead naming the work and the exception you think applies, and await the lead's answer before proceeding.

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;
}

Test runner. The commands below are illustrative. Use whatever test runner the BC's own harness uses. For a Python BC this is typically pytest <path>; a TypeScript BC would use npm test <path>. Substitute your BC's runner throughout — the discipline is the same regardless of runner.

RED - Write Failing Test

Write one minimal test showing what should happen.

```python def test_retries_failed_operations_3_times(): attempts = {"count": 0}

def operation():
    attempts["count"] += 1
    if attempts["count"] < 3:
        raise RuntimeError("fail")
    return "success"

result = retry_operation(operation)

assert result == "success"
assert attempts["count"] == 3

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

<Bad>
```python
def test_retry_works(mocker):
    mock = mocker.Mock(side_effect=[RuntimeError(), RuntimeError(), "success"])
    retry_operation(mock)
    assert mock.call_count == 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.

pytest path/to/test_retry.py   # or your BC's runner

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.

```python def retry_operation(fn): for i in range(3): try: return fn() except Exception: if i == 2: raise raise RuntimeError("unreachable") ``` Just enough to pass ```python def retry_operation(fn, max_retries=None, backoff=None, on_retry=None): # YAGNI ... ``` Over-engineered

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

Verify GREEN - Watch It Pass

MANDATORY.

pytest path/to/test_retry.py   # or your BC's runner

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 behavior. (Each behavior is a bd sub-issue when the work decomposes — see The Two Loops.)

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

def test_rejects_empty_email():
    result = submit_form({"email": ""})
    assert result["error"] == "Email required"

Verify RED

$ pytest   # or your BC's runner
FAILED: KeyError: 'error'  (expected 'Email required', got nothing)

GREEN

def submit_form(data):
    if not (data.get("email") or "").strip():
        return {"error": "Email required"}
    # ...

Verify GREEN

$ pytest   # or your BC's runner
PASSED

REFACTOR Extract validation for multiple fields if needed.

Verification Checklist

Before marking work complete (emitting work_done):

• The assigned Gherkin scenario(s) pass (the outer loop — this is what work_done proves)
• The working tree is clean (no uncommitted changes, no stray files)
• 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

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

What work_done surfaces. work_done evidence is, and remains, the assigned scenario(s) pass against a clean working tree. The TDD discipline in this skill is the engineering process you follow to get there — at this time it is not evidence the BC surfaces in work_done. Run the inner loop because it produces correct, trustworthy code; do not attempt to report RED-GREEN-REFACTOR transcripts, per-test fail-watch logs, or TDD adherence as part of the work_done payload.

When Stuck

Problem	Solution
Don't know how to test	Write wished-for API. Write assertion first. Still stuck? Emit a clarify to the lead.
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? Write failing test reproducing it. Follow TDD cycle. Test proves fix and prevents regression.

Never fix bugs without a test.

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 lead's decision. If you believe an exception applies, emit a clarify to the lead and await the answer — never grant yourself the exception.