green-phase

Skill Type: Actuator (TDD Workflow) Purpose: Write minimal code to make failing tests pass (GREEN phase) Prerequisites:

• Tests exist and are FAILING (from red-phase)
• Requirement key (REQ-*) with details

---

Agent Instructions

You are in the GREEN phase of TDD (RED → GREEN → REFACTOR).

Your goal is to write MINIMAL code to make the failing tests pass.

Key principle: Write the simplest code that works. Do NOT over-engineer. Refactoring comes later.

---

Workflow

Step 1: Review Failing Tests

Read the test file to understand:

• What functionality is being tested?
• What are the expected inputs and outputs?
• What business rules must be enforced?

Example:

# tests/auth/test_login.py shows we need:
- login(email: str, password: str) -> LoginResult
- LoginResult with fields: success, user, error
- Email validation (BR-001)
- Password length validation (BR-002)
- Account lockout logic (BR-003)

---

Step 2: Determine Implementation File Location

Follow project conventions:

Python:

tests/auth/test_login.py  → src/auth/login.py
tests/services/test_payment.py → src/services/payment.py

TypeScript:

src/auth/login.test.ts    → src/auth/login.ts
src/services/payment.test.ts → src/services/payment.ts

Java:

src/test/java/auth/LoginTest.java → src/main/java/auth/Login.java

---

Step 3: Implement Minimal Code

Write the simplest code that makes tests pass:

# src/auth/login.py

# Implements: <REQ-ID>
# Business Rules: BR-001, BR-002, BR-003

import re
from dataclasses import dataclass
from typing import Optional
from datetime import datetime, timedelta

@dataclass
class LoginResult:
    success: bool
    user: Optional['User'] = None
    error: Optional[str] = None

class User:
    def __init__(self, email: str, password_hash: str):
        self.email = email
        self.password_hash = password_hash
        self.failed_attempts = 0
        self.locked_until: Optional[datetime] = None

    def check_password(self, password: str) -> bool:
        # Simplified: In real code, use bcrypt
        return self.password_hash == hash_password(password)

# Implements: <REQ-ID>
def login(email: str, password: str) -> LoginResult:
    # Implements: BR-001 (email validation)
    if not validate_email(email):
        return LoginResult(success=False, error="Invalid email format")

    # Implements: BR-002 (password minimum length)
    if len(password) < 12:
        return LoginResult(success=False, error="Password must be at least 12 characters")

    # Get user from database (simplified)
    user = get_user_by_email(email)
    if not user:
        return LoginResult(success=False, error="User not found")

    # Implements: BR-003 (account lockout)
    if user.locked_until and datetime.now() < user.locked_until:
        return LoginResult(success=False, error="Account locked. Try again in 15 minutes")

    # Check password
    if not user.check_password(password):
        user.failed_attempts += 1

        # Implements: BR-003 (lock after 3 attempts)
        if user.failed_attempts >= 3:
            user.locked_until = datetime.now() + timedelta(minutes=15)

        return LoginResult(success=False, error="Invalid password")

    # Success - reset failed attempts
    user.failed_attempts = 0
    user.locked_until = None

    return LoginResult(success=True, user=user)

# Implements: BR-001 (email validation)
def validate_email(email: str) -> bool:
    pattern = r'^[a-zA-Z0-9._%+-]+@[a-zA-Z0-9.-]+\.[a-zA-Z]{2,}$'
    return re.match(pattern, email) is not None

def hash_password(password: str) -> str:
    # Simplified: In real code, use bcrypt
    return str(hash(password))

def get_user_by_email(email: str) -> Optional[User]:
    # Simplified: In real code, query database
    # For now, return mock user for testing
    return User(email, hash_password("SecurePass123!"))

Key implementation principles:

• ✅ Tag code with REQ-* keys in comments
• ✅ Tag business rule implementations with BR-* keys
• ✅ Write just enough code to pass tests (no gold-plating)
• ✅ Use clear, descriptive names
• ✅ Keep functions focused (single responsibility)
• ✅ Don't worry about perfect code (refactoring comes next)

---

Step 4: Run Tests (Expect SUCCESS)

Run the test suite:

# Python
pytest tests/auth/test_login.py -v

# TypeScript/JavaScript
npm test src/auth/login.test.ts

# Java
mvn test -Dtest=LoginTest

Expected output:

tests/auth/test_login.py::test_login_with_valid_credentials PASSED ✓
tests/auth/test_login.py::test_login_fails_with_invalid_email PASSED ✓
tests/auth/test_login.py::test_login_fails_with_short_password PASSED ✓
tests/auth/test_login.py::test_account_locked_after_3_failed_attempts PASSED ✓

4 tests passed in 0.12s

✅ All tests PASSING! This is what we want in GREEN phase.

⚠️ If tests still FAIL: Fix implementation and retry until all tests pass.

---

Step 5: Verify Test Coverage

Check coverage:

# Python
pytest --cov=src/auth tests/auth/test_login.py --cov-report=term-missing

# TypeScript
npm test -- --coverage

# Java
mvn test jacoco:report

Expected: Coverage should be high (aim for 80%+ overall, 100% for critical paths).

---

Step 6: Commit Implementation (GREEN Commit)

Create commit with implementation:

git add src/auth/login.py
git commit -m "GREEN: Implement <REQ-ID>

Implement user login functionality with email/password.

Implements:
- <REQ-ID>: User login
- BR-001: Email validation (regex pattern)
- BR-002: Password minimum 12 characters
- BR-003: Account lockout after 3 failed attempts (15min)

Tests: 4 tests passing ✓
Coverage: 95%
"

Commit message format:

• Prefix: GREEN:
• Brief description
• List of REQ-* and BR-* implemented
• Test status (all passing)

---

Output Format

When you complete the GREEN phase, show:

[GREEN Phase - <REQ-ID>]

Implementation: src/auth/login.py

Code Created:
  ✓ LoginResult dataclass
  ✓ User class
  ✓ login() function (<REQ-ID>)
  ✓ validate_email() function (BR-001)
  ✓ hash_password() helper
  ✓ get_user_by_email() helper

Business Rules Implemented:
  ✓ BR-001: Email validation (regex)
  ✓ BR-002: Password minimum 12 characters
  ✓ BR-003: Account lockout after 3 attempts

Running tests...
  tests/auth/test_login.py::test_login_with_valid_credentials PASSED ✓
  tests/auth/test_login.py::test_login_fails_with_invalid_email PASSED ✓
  tests/auth/test_login.py::test_login_fails_with_short_password PASSED ✓
  tests/auth/test_login.py::test_account_locked_after_3_failed_attempts PASSED ✓

Result: 4/4 tests PASSING ✓ (GREEN phase)

Coverage: 95% (38/40 lines covered)

Commit: GREEN: Implement <REQ-ID>

✅ GREEN Phase Complete!
   Next: Invoke refactor-phase skill to improve code quality

---

Prerequisites Check

Before invoking this skill, ensure:

1. Tests exist and are FAILING (from red-phase)
2. Requirement details available (what to implement)

If prerequisites not met:

• No tests → Invoke red-phase skill first
• Tests already passing → Already implemented, skip to refactor-phase

---

Next Steps

After GREEN phase completes:

1. Do NOT refactor yet (tests must pass first)
2. Invoke refactor-phase skill to improve code quality and eliminate tech debt
3. Tests should STILL PASS after refactoring

---

Implementation Strategies

Strategy 1: Simplest Thing That Works

# First implementation (naive, but works)
def login(email: str, password: str) -> LoginResult:
    if email == "user@example.com" and password == "SecurePass123!":
        return LoginResult(success=True)
    return LoginResult(success=False)

Then improve to handle more cases until all tests pass.

Strategy 2: Test-by-Test

Implement code to pass one test at a time:

1. Make test_login_with_valid_credentials pass
2. Make test_login_fails_with_invalid_email pass
3. Make test_login_fails_with_short_password pass
4. Make test_account_locked_after_3_failed_attempts pass

Strategy 3: Fake It Till You Make It

Start with hard-coded values, then generalize:

# Step 1: Hard-coded (passes one test)
def validate_email(email: str) -> bool:
    return email == "user@example.com"

# Step 2: Generalize (passes all tests)
def validate_email(email: str) -> bool:
    pattern = r'^[a-zA-Z0-9._%+-]+@...'
    return re.match(pattern, email) is not None

---

Common Pitfalls to Avoid

❌ Over-engineering: Don't add features not tested ❌ Premature optimization: Don't optimize before refactor phase ❌ Perfect code: Don't worry about code quality yet (refactor phase handles this) ❌ Skipping tests: All tests must pass before moving to refactor ❌ Adding untested code: Every line should be tested

✅ Do this instead:

• Write minimal code
• Make tests pass
• Move to refactor phase
• Improve code quality there

---

Notes

Why minimal implementation?

• Prevents over-engineering (YAGNI - You Aren't Gonna Need It)
• Forces focus on requirements (only implement what's tested)
• Faster to refactor simple code
• Easier to understand

GREEN phase mantra: "Make it work, then make it right"

• GREEN phase = make it work
• REFACTOR phase = make it right

Homeostasis Goal:

desired_state:
  tests_passing: true
  all_requirements_implemented: true
  code_coverage: >= 80%

"Excellence or nothing" 🔥