Test Runner Skill

Run the project's test suite, analyze results, and optionally fix failing tests.

Usage

/test-runner                    # Run all tests
/test-runner src/auth/          # Run tests for a specific directory
/test-runner --fix              # Run tests and fix failures
/test-runner --coverage         # Run tests with coverage report

Steps

1. •

   Detect the test framework Check the project for:

   • •jest.config.* or "jest" in package.json → Jest
   • •vitest.config.* → Vitest
   • •pytest.ini, pyproject.toml [tool.pytest] → pytest
   • •go test → Go testing
   • •cargo test → Rust testing
   • •.rspec → RSpec
   • •phpunit.xml → PHPUnit
2. •

   Build the test command

   • •All tests: use the project's test script (npm test, pytest, etc.)
   • •Specific path: pass the path to the test runner
   • •Coverage: add the appropriate coverage flag
   • •Use verbose output when available for better diagnostics
3. •

   Run the tests Execute the test command and capture the output.

4. •

   Analyze the results Parse the output to identify:

   • •Total tests, passed, failed, skipped
   • •Which specific tests failed
   • •Error messages and stack traces
   • •Coverage percentages (if requested)
5. •

   Report results

   code

   ## Test Results
   
   ✓ X passed | ✗ X failed | ○ X skipped | Total: X
   
   ### Failures
   - test_name (file:line) — Error: description
   
   ### Coverage (if requested)
   - Statements: X%
   - Branches: X%
   - Functions: X%
   - Lines: X%
   

6. •

   Fix failures (if --fix flag) For each failing test:

   • •Read the test file and the source file it tests
   • •Determine if the issue is in the test or the source code
   • •If the test is wrong: fix the test
   • •If the source is wrong: fix the source
   • •Re-run to verify the fix
   • •Repeat until all tests pass

Rules

• •Always run the existing test command — don't invent your own
• •When fixing, prefer fixing source code bugs over changing test expectations
• •Never delete or skip failing tests to make the suite pass
• •If a test is genuinely outdated, explain why before updating it
• •Show the full error output for failures so the user can review