Java unit test generator

Java Unit Test Generator

Instructions

Step 1: Scope Definition

Before generating tests, ask:

1. Which class/method?
2. Coverage goal? (Basic = happy path | Standard = + errors | Comprehensive = + edge cases)
3. Framework? (JUnit 4, JUnit 5, TestNG)
4. Mocking? (Mockito, EasyMock, none)

Provide an upfront estimate: 10-15 credits per method.

Step 2: Generate 3 Tests Per Batch

For each method, generate exactly 3 tests:

1. Happy path
2. Error handling
3. Edge case

Use this structure:

@Test
@DisplayName("Should [expected behavior] when [condition]")
void test[MethodName][Scenario]() {
    // Given - setup
    // When - execute
    // Then - verify
}

STOP after each batch. Ask the user to run and confirm before continuing.

Step 3: Validation Gate

Do NOT generate more tests until the user confirms the current batch works.

Validation command:

python scripts/validate_tests.py <TestClassName>

Step 4: Fix or Skip (Max 2 Attempts)

If tests fail:

• Attempt 1: Fix compilation errors (imports, syntax)
• Attempt 2: Fix logic errors (assertions, test data)
• After 2 failures: Skip the test, document the issue, and move on

When skipping, say:

"This test failed twice. Skipping to preserve credits. Issue: [describe]. Options: (1) skip, (2) provide more context, (3) simplify scope."

Never attempt the same fix more than twice. Mark flaky tests @Disabled.

Step 5: Iterate or Finish

After a passing batch, offer:

• 3 more tests for the same method
• Move to next method
• Stop and produce summary

Step 6: Summary Report

Methods Tested: N
Total Tests: N
Skipped: N (with reasons)
Coverage: happy path / error / edge case breakdown

Measuring Credit Cost for Your Code

Credit costs vary by Kiro account, model, and code complexity. To get accurate numbers for your environment, follow these steps.

1. Establish your baseline (one-time)

Run the built-in benchmark with a simple sample class:

python track_credits.py benchmark java-simple

This measures the cheapest possible unit test on your account (e.g., 0.063 credits/test).

2. Estimate before running on your code

Your real classes are more complex. The tool scores complexity based on:

What makes tests cost more	How it's scored
More lines of code	+1 for every 50 lines
More mock dependencies	+1 for each mock
More branching (if/else/switch)	+1 for every 3 branches
More framework annotations	+1 for every 3 annotations

python track_credits.py estimate \
  --type java-unit \
  --source-lines 150 \
  --mocks 3 \
  --branches 6 \
  --annotations 4 \
  --tests 3

3. Measure actual cost after running

# Note credit balance before and after, then record:
python track_credits.py add "MyService-3tests" <before> <after> "3 JUnit5 tests, Mockito"

Your measurements improve future estimates automatically. See ../../../skills-workshop/credit-estimation/CREDIT-ESTIMATION-GUIDE.md for the full guide.

References

• references/junit5-patterns.md - Common JUnit 5 patterns
• references/examples/UserServiceTest.java - Complete example
• ../../../skills-workshop/credit-estimation/CREDIT-ESTIMATION-GUIDE.md - Credit measurement methodology
• JUnit 5 User Guide
• Mockito Documentation