When to Apply This Skill
- A developer provides source code and asks for unit tests
- A CI/CD pipeline triggers test generation for a new commit or PR
- A team wants to bootstrap tests for an untested module
- Any input (code, spec, requirement) needs to produce a runnable test file
Atomic Skills to Load First
Read these files before executing any step — they contain the detailed instructions this orchestration skill depends on:
../unit-test-generating-unit-tests/SKILL.md— Test matrix methodology, AAA pattern, mocking, parametrization, exception testing, and framework-specific code examples../unit-test-analyzing-code-coverage/SKILL.md— Coverage type definitions and gap report format, used when producing coverage_plan.md at the end of generation
Execution Steps
Read references/index.md before executing any step.
Step 1 — Classify Input
Identify what has been provided and how to interpret it:
source_code → extract signatures, logic paths, side effects requirements → extract acceptance criteria and business rules openapi_spec → extract endpoints, parameters, response schemas mixed → combine all of the above
Also identify: target language, test framework (or auto-detect), scope (function | class | module).
Step 2 — Invoke test-unit-gen-agent
Hand off to ../../agents/unit-testing/test-unit-gen-agent.md with the classified input.
The agent will:
- Build the test case matrix (following unit-test-generating-unit-tests skill)
- Author all test cases using AAA pattern
- Assemble the test file and coverage_plan.md
Step 3 — Invoke test-unit-review-agent
Hand off the generated suite to ../../agents/unit-testing/test-unit-review-agent.md.
The agent returns one of three verdicts: APPROVED → proceed to delivery APPROVED_WITH_NOTES → proceed to delivery with report attached REQUIRES_REVISION → return to Step 2 with annotated fix list
Maximum revision cycles: 2. If still REQUIRES_REVISION after 2 cycles, deliver with full issue report for human resolution.
Output Deliverables
test_<module>.<ext> <- runnable test suite (pytest, Jest, JUnit 5, xUnit, etc.)
coverage_plan.md <- coverage intent summary and gap notes
review_report.md <- quality verdict and any reviewer notes