Test Skill

Runs API integration tests and the full Playwright E2E suite.

Step 1 — API integration tests

dotnet test --project api/tests/Skoleoverblikket.Api.IntegrationTests/Skoleoverblikket.Api.IntegrationTests.csproj --configuration Release

Runs all tUnit integration tests via Testcontainers. Needs Docker running. If Docker is not running, say so explicitly rather than claiming failure is a code bug.

Step 2 — Playwright E2E

cd web && npx playwright install chromium && npx playwright test --reporter=line

playwright install is a no-op when the browser is current; it's there because a @playwright/test bump otherwise fails global setup with "Executable doesn't exist".

Playwright config starts the full Aspire stack automatically via aspire run --non-interactive if not already up. Pass SKIP_ASPIRE=1 only if the stack is already running.

Do NOT pre-start Aspire manually. Do NOT poll for port readiness. Just run the command — Playwright handles it.

What to report

After both steps, report clearly:

• API tests: pass/fail count and any failure details (test name, assertion error, file:line)
• E2E tests: pass/fail count, which tests failed (name and file, verbatim), failure details
• Whether failure is code or environment — if a test fails because Aspire isn't running or Docker is down, say so explicitly

Fixing failures

If tests fail:

1. Read the failure output carefully before touching any code
2. Check whether the failure is environmental (stack not running, wrong port, Docker down, leftover state) vs a real regression
3. If it's a real regression, investigate the root cause in the relevant test file and the frontend/API code it exercises
4. Fix the underlying issue, then re-run to confirm green

Do not comment out failing tests or lower assertion thresholds to make tests pass.

Test locations

• API integration tests: api/tests/
• E2E spec files: web/tests/e2e/
• Playwright config: web/playwright.config.ts