All Weeks
DAY 29

Prompt Engineering

LLMs as force multipliers. Stay in control while accelerating development.
  1. Build reusable prompt templates for POM generation
  2. Use AI to generate RecruitmentPage.ts locators
  3. Implement a "review mindset" workflow for AI code
  4. Document context engineering tricks that improve accuracy
DAY 30

Self-Healing Tests

Tests that survive UI changes by falling back to alternative locators.
  • Build a HealingLocator utility with fallback chains
  • Integrate AI API calls for last-resort locator suggestions
  • Simulate UI refactors and measure rescue rate
  • Define team policies for bridge-healing vs permanent fixing
DAY 31

Testing AI Apps

Validate non-deterministic UIs. Test intent, not just exact text.
  1. Test chatbot responses with semantic matching
  2. Verify AI response quality (length, JSON structure)
  3. Implement "LLM-as-judge" for output validation
DAY 32

AI Agents & MCP

Connect your browser to LLMs via Model Context Protocol.
  1. Install npx @playwright/mcp@latest
  2. Connect Claude/Cursor to your live browser session
  3. Use agents to generate test plans from plain requirements
  4. Have the agent "heal" a broken test in real-time
DAY 33

Vibe Testing

AI-augmented exploratory testing. Discover the "unknown unknowns".
  • Ask AI for "attack ideas" on a new feature
  • Record sessions with --trace on
  • Convert findings into regression tests
DAY 34

Cloud Execution

True cross-browser and real-device coverage at enterprise scale.
  • Connect to LambdaTest or BrowserStack via CDP
  • Run tests on real mobile devices (iPhone/Pixel)
  • Consolidate cloud results into your local Report Portal
  • Analyze cost vs coverage for your automation strategy
DAY 35

Foundation: SuiteCRM

New Application. Same Framework. Prove your skills transfer.
  1. Map SuiteCRM login and Accounts modules
  2. Setup SuiteCRM project in playwright.config.ts
  3. Write globalSetup for SuiteCRM auth
Previous Day Next Day