---

name: quality-e2e-testing description: > End-to-end testing skill using Playwright. Includes advanced mocking, network manipulation, and chaos engineering. version: "2.0.0" author: "j4flmao" license: "MIT" type: skill compatibility: claude-code: true cursor: true codex: true windsurf: true tags:

• testing
• playwright
• e2e

---

Quality E2E Testing

Purpose

Comprehensive end-to-end testing capabilities for robust applications. Includes Playwright, chaos engineering, and more.

Core Principles

1. Determinism over speed.
2. Isolate test states.
3. Network layer is fully controlled.
4. Assume nothing, test everything.
5. Chaos engineering uncovers hidden limits.

Agent Protocol

• Triggers: e2e test request
• Input Context Required: application url, auth state
• Output Artifact: test report
• Response Formats:

{ "status": "pass", "coverage": "90%" }

Decision Matrix

[E2E Request]
    |
    v
 Is local? ---> (Yes) -> Run headless Playwright
    |
   (No) -> Run in container

Detailed Architectural Overview

[Browser] <--> [Playwright] <--> [Test Runner]

Lifecycle:
Init -> Setup State -> Execute -> Teardown

Workflow Steps

Phase 1: Setup

1. Init browser
2. Load state
3. Route mock
4. Wait ready

Phase 2: Execution

1. Navigate
2. Interact
3. Assert
4. Cleanup

Phase 3: Chaos

1. Drop packets
2. Throttle CPU
3. Assert recovery
4. Analyze

Phase 4: Reporting

1. Collect traces
2. Generate HTML
3. Upload artifacts
4. Alert

Phase 5: Cleanup

1. Close browser
2. Clean state
3. Notify
4. End

Phase 6: Analysis

1. Parse logs
2. Find flakiness
3. Update metrics
4. Store

Extended Troubleshooting Guide

Symptom	Primary Cause	Mitigation Action
Timeout	Network slow	Increase timeout
Flaky	Async race	Use waitForSelector
No Auth	Bad token	Refresh session
Crash	OOM	Increase memory limit
Element not found	DOM changed	Update locators
Blank page	JS error	Check console logs

Complete Execution Scenario

Start -> Mock -> Test -> Success -> End

Rules and Guidelines

1. No arbitrary waits.
2. Always mock external APIs.
3. Keep tests isolated.
4. Clean up after each test.
5. Use page object model.

Reference Guides

• Playwright Mocking
• Chaos Engineering
• Network Throttling
• Flaky Test Mitigation
• Visual Regression
• A11Y Testing
• Test Data
• CI CD

Handoff

See ci-integration for pipelines.