home / skills / d-oit / do-novelist-ai / e2e-test-optimizer
This skill optimizes Playwright E2E tests by removing waitForTimeout anti-patterns, enabling test sharding, and speeding CI while improving reliability.
npx playbooks add skill d-oit/do-novelist-ai --skill e2e-test-optimizerReview the files below or copy the command above to add this skill to your agents.
---
name: e2e-test-optimizer
description:
Optimizes Playwright E2E tests by removing waitForTimeout anti-patterns,
implementing smart waits, and enabling test sharding for parallel execution.
Use when tests timeout, CI execution exceeds limits, or test reliability
issues occur.
---
# E2E Test Optimizer
## Quick Start
This skill helps fix three critical E2E test issues:
1. **Anti-pattern removal**: Replace arbitrary `waitForTimeout` with smart waits
2. **Test sharding**: Enable parallel test execution in CI
3. **Mock optimization**: Reduce mock setup overhead
### When to Use
- E2E tests timing out or failing due to arbitrary waits
- CI execution time exceeds 10 minutes
- Need test parallelization for faster feedback
- Test flakiness from race conditions
## Smart Wait Patterns
### Navigation Waits
```typescript
// ❌ Arbitrary wait
await page.click('[data-testid="nav-dashboard"]');
await page.waitForTimeout(1000);
// ✅ Wait for specific element
await page.click('[data-testid="nav-dashboard"]');
await expect(page.getByTestId('dashboard-content')).toBeVisible({
timeout: 5000,
});
```
### State Transition Waits
```typescript
// ❌ Arbitrary wait after action
await page.getByRole('button', { name: 'Generate' }).click();
await page.waitForTimeout(2000);
// ✅ Wait for loading state completion
await page.getByRole('button', { name: 'Generate' }).click();
await expect(page.getByTestId('loading-spinner')).not.toBeVisible();
await expect(page.getByTestId('result')).toBeVisible();
```
### Network Waits
```typescript
// ❌ Arbitrary wait after submit
await page.getByRole('button', { name: 'Save' }).click();
await page.waitForTimeout(1500);
// ✅ Wait for network idle or success message
await page.getByRole('button', { name: 'Save' }).click();
await page.waitForLoadState('networkidle');
// OR
await expect(page.getByText('Saved successfully')).toBeVisible();
```
## Test Sharding Setup
Add to `.github/workflows/ci.yml`:
```yaml
e2e-tests:
name: 🧪 E2E Tests [Shard ${{ matrix.shard }}/3]
runs-on: ubuntu-latest
timeout-minutes: 30
needs: build-and-test
strategy:
fail-fast: false
matrix:
shard: [1, 2, 3]
steps:
- name: Run Playwright tests
run: pnpm exec playwright test --shard=${{ matrix.shard }}/3
env:
CI: true
```
**Expected improvement**: 60-65% faster (27 minutes → 9-10 minutes)
## Detection Commands
```bash
# Find all waitForTimeout usage
grep -r "waitForTimeout" tests/specs/*.spec.ts
# Count per file
grep -c "waitForTimeout" tests/specs/*.spec.ts
```
## Standard Test Structure
```typescript
import { test, expect } from '@playwright/test';
import { setupGeminiMock } from '../utils/mock-ai-gateway';
test.describe('Feature Name', () => {
test.beforeEach(async ({ page }) => {
await setupGeminiMock(page);
await page.goto('/');
await page.waitForLoadState('networkidle');
});
test('should perform action', async ({ page }) => {
// Navigate
await page.getByTestId('nav-target').click();
await expect(page.getByTestId('target-page')).toBeVisible();
// Interact
await page.getByRole('button', { name: 'Action' }).click();
await expect(page.getByTestId('loading')).not.toBeVisible();
// Assert
await expect(page.getByTestId('result')).toBeVisible();
});
});
```
## Element Selection Priority
```typescript
// ✅ Best: data-testid (stable, explicit)
await page.getByTestId('project-card');
// ✅ Good: role + name (semantic, accessible)
await page.getByRole('button', { name: 'Create' });
// ⚠️ OK: text (can break with i18n)
await page.getByText('Dashboard');
// ❌ Avoid: CSS selectors (brittle)
await page.locator('.project-card > div.title');
```
## Optimization Workflow
### Phase 1: Analysis
1. Scan test files for `waitForTimeout`
2. Count occurrences per file
3. Prioritize files by occurrence count
### Phase 2: Fix Anti-Patterns
1. Start with highest-priority file
2. Replace each `waitForTimeout` with smart wait
3. Run tests after each file: `playwright test path/to/file.spec.ts`
4. Commit per-file changes
### Phase 3: Implement Sharding
1. Update CI workflow with matrix strategy
2. Test locally: `playwright test --shard=1/3`
3. Push and monitor all 3 shards in CI
### Phase 4: Validation
1. Verify all tests pass
2. Confirm execution time < 10 minutes
3. Check shard balance (±2 min variance acceptable)
## Common Issues
**Tests still timing out after fix**
- Increase timeout on slow operations:
`expect(...).toBeVisible({ timeout: 10000 })`
**Unbalanced shards (one takes much longer)**
- Use `--grep` to manually distribute heavy tests across shards
**Mock setup still slow**
- Implement global browser warm-up (see MOCK-OPTIMIZATION-GUIDE.md in
tests/docs)
## Success Criteria
- Zero `waitForTimeout` calls in active tests
- CI execution time < 10 minutes
- All shards complete within ±2 minutes of each other
- 100% test pass rate (no flaky tests)
## References
See tests/docs/ for detailed guides:
- MOCK-OPTIMIZATION-GUIDE.md - Mock performance patterns
- MOCK-PERFORMANCE-ANALYSIS.md - Optimization results
External documentation:
- Playwright Best Practices: https://playwright.dev/docs/best-practices
- Test Sharding: https://playwright.dev/docs/test-sharding
This skill optimizes Playwright end-to-end tests by removing waitForTimeout anti-patterns, introducing smart waits, and enabling test sharding for parallel CI execution. It reduces flakiness, shortens CI runtime, and standardizes element selection and test structure. The goal is reliable tests that complete within CI time limits.
The tool scans test files for waitForTimeout occurrences, prioritizes files by frequency, and replaces arbitrary sleeps with smart waits (visibility, networkidle, or specific state checks). It provides patterns for navigation, state transitions, and network waits, and offers a recipe to add GitHub Actions matrix-based sharding for parallel execution. A recommended workflow guides per-file fixes, local validation, CI sharding setup, and performance validation.
Will replacing waitForTimeout break tests on slower CI runners?
No—smart waits explicitly wait for elements, states, or network idle and include adjustable timeouts so tests are resilient across environments.
How do I distribute tests when shards are unbalanced?
Use --grep to move heavy tests manually, or split files with long-running scenarios across different shards to balance execution time.