ワンクリックで
unit-test-executor
Allows to execute unit tests and systematically resolve all test failures.
Codex または Claude でインストール この Prompt をコピーして Codex、Claude、または他のアシスタントに貼り付けると、Skill ページを確認してインストールできます。
メニュー
Allows to execute unit tests and systematically resolve all test failures.
Codex または Claude でインストール この Prompt をコピーして Codex、Claude、または他のアシスタントに貼り付けると、Skill ページを確認してインストールできます。
SOC 職業分類に基づく
Allows to create new AI agents following the project's architecture patterns.
Allows to review an open GitHub PR for the current branch, applying project-specific and Java best practice criteria, and adds inline comments directly to the PR.
Allows to create new unit tests following established patterns and conventions.
Allows to commit all files and push them to the remote repository.
Allows to prepares the current branch for a pull request. Use when ready to create a PR or before committing changes.
Allows to plan and guide the implementation of new features or modification of existing features, ensuring best practices, modern design patterns, and proper planning. Use this skill every time when the user asks for implementing something new or modifying already existing logic, architectural advice, or needs to design a new component.
| name | unit-test-executor |
| description | Allows to execute unit tests and systematically resolve all test failures. |
// turbo-all
This skill provides a systematic workflow for running the test suite, diagnosing failures, and resolving them.
Note: For writing new tests or understanding test patterns, see the Unit Testing skill.
Analyze the parent POM file and the POM file of each module which tests need to be executed. Extract the versions of JUnit, Mockito, AssertJ. Remember those versions because they will be needed in the next tasks.
Follow this systematic process when identifying and resolving test failures.
Run all tests to evaluate the current health of the codebase.
mvn test
Analyze the console output or target/surefire-reports to categorize failures:
| Category | Description | Priority |
|---|---|---|
| Compilation Errors | Code doesn't compile | Highest |
| Runtime Exceptions | NPE, IndexOutOfBounds, ClassCast, etc. | High |
| Assertion Failures | Expected vs Actual mismatch | Medium |
| Mockito Errors | UnnecessaryStubbingException, PotentialStubbingProblem | Medium |
For each failure, prioritize low-level/unit failures.
Reproduce: Run only the failing test class to isolate the issue.
mvn test -Dtest=ClassName
Debug:
lenient() might be needed if the stub is conditional).MockitoAnnotations.openMocks(this)).Analyze:
Fix:
Verify: Run the specific test again.
mvn test -Dtest=ClassName
Repeat: Move to the next failing test class.
Once individual issues are resolved, run the full module suite to ensure no side effects.
mvn test -pl module_name
If new failures appear, loop back to Step 3 for those tests.
Run the complete project test suite before finishing the task.
mvn test
All tests must pass before the task is considered complete.
| Goal | Command |
|---|---|
| Run all tests | mvn test |
| Run tests in a module | mvn test -pl module_name |
| Run a specific test class | mvn test -Dtest=ClassNam e |
| Run a specific test method | mvn test -Dtest=ClassNam e#methodName |
| Run tests with output | mvn test -Dtest=ClassNam e -X |
| Skip tests during build | mvn install -DskipTests |
SafeToAutoRun to true for all validation and test execution commands (e.g., mvn test).>) to capture command output. Instead, rely on the agent's built-in command output capturing.