Backend Test Writer

Generate comprehensive backend tests for MERN stack code. Analyzes file type, detects project conventions, produces tests that actually run.

Philosophy: Smart defaults, zero config. Detect everything from project.

<workflow>

Workflow

Copy and track progress:

• [ ] Phase 0: Infrastructure check
• [ ] Phase 1: Analyze target files
• [ ] Phase 2: Generate tests
• [ ] Phase 3: Report summary

Phase 0: Infrastructure Check

Before generating tests, verify setup:

1. Check package.json for test framework (Jest/Vitest/Mocha)
2. If none → prompt: "Set up Jest + Supertest?"
3. Check for mongodb-memory-server (needed for integration tests)
4. Detect test file convention (colocated vs __tests__/ vs tests/)

Stop if: No test framework and user declines setup.

Phase 1: Analyze Target Files

Pattern	Type	Test Approach
routes/, *.routes.js	Route	Integration (Supertest + real DB)
controllers/	Controller	Integration
services/	Service	Unit (mocked deps)
models/, *.model.js	Model	Unit (validation tests)
middleware/	Middleware	Unit (mock req/res/next)
utils/, helpers/	Utility	Unit (pure functions)

Override: User can specify --unit or --integration.

Phase 2: Generate Tests

Process files sequentially with progress. User can stop anytime.

Each test includes:

• Proper imports for detected framework
• Setup/teardown (DB connection, cleanup)
• Comprehensive coverage:
  • Success cases (happy path)
  • Validation errors (400)
  • Not found (404)
  • Auth failures (401/403) if protected
  • Edge cases (duplicates, empty, null)

Reference: See test-patterns.md for complete code examples.

Phase 3: Report

Generated: X test files
Coverage: Y test cases total
Next: Run `npm test` to verify

</workflow>

<quick-reference>

Quick Reference

File Type	Imports	DB Setup
Route	supertest, mongodb-memory-server	Real (in-memory)
Service	jest	Mocked
Model	mongoose	Mocked
Middleware	jest	None

Test Structure Pattern

describe('[Resource] [Method]', () => {
  describe('success cases', () => {
    it('should [expected behavior]', async () => {});
  });
  describe('validation errors', () => {
    it('should return 400 for [invalid case]', async () => {});
  });
  describe('edge cases', () => {
    it('should handle [edge case]', async () => {});
  });
});

</quick-reference>

<checklists>

Checklists

Infrastructure (Check First)

• [ ] Test framework in package.json
• [ ] Test script defined ("test": "jest")
• [ ] Supertest installed (integration tests)
• [ ] mongodb-memory-server installed (DB tests)

Per-File Generation

• [ ] Check for existing tests first (gap analysis if found)
• [ ] Correct imports for file type
• [ ] Setup/teardown included
• [ ] Happy path tested
• [ ] Error cases tested (400, 404, 401)
• [ ] Edge cases tested (domain-specific: date→DST/timezone, money→precision, etc.)
• [ ] No secrets in assertions
• [ ] Async/await handled properly
• [ ] Priority assigned to each test (P0=critical, P1=important, P2=nice-to-have)
• [ ] Test helper factories suggested for complex inputs

</checklists>

<common-mistakes>

Common Mistakes

Mistake	Fix
Starting server in tests	Import app, let Supertest handle it
No DB cleanup	Add afterEach with deleteMany({})
Testing implementation	Test behavior through HTTP interface
Missing async/await	Await async operations
Mocking in integration tests	Use real DB for integration

</common-mistakes>

<guidelines>

Guidelines

• Don't generate tests for unread code
• Don't skip infrastructure check
• Don't generate only happy paths
• Don't forget cleanup between tests

</guidelines>

<references>

Reference Files

Load when implementing specific patterns:

When	Reference
Writing any test	test-patterns.md
Setting up test infrastructure	test-setup.md

</references>