Skip to content

task_plan.md

Latest commit

 

History

History
42 lines (34 loc) · 1.8 KB

task_plan.md

File metadata and controls

42 lines (34 loc) · 1.8 KB
Raw

Task Plan - Selenium to Playwright Converter

Project Phases

🟢 Phase 0: Initialization (Current)

  • Initialize project memory files (task_plan.md, findings.md, progress.md, gemini.md)
  • Discovery Questions for the User
  • Define Data Schema in gemini.md

🏗️ Phase 1: B - Blueprint (Vision & Logic)

  • Research existing Selenium to Playwright conversion tools/patterns
  • Define conversion mapping (Selenium Java -> Playwright TS/JS)
  • Define System Architecture:
    • Frontend: HTML/JS dashboard (due to local environment constraints)
    • Backend: PowerShell-based conversion scripts interacting with Ollama.
    • Logic: Prompt-engineered conversion using llama3.2:1b (pending) or tinyllama.
  • Finalize Blueprint

⚡ Phase 2: L - Link (Connectivity)

  • Setup local environment (Ollama on port 11434)
  • Verify Ollama accessibility and model availability
  • [wip] Pull optimized model (llama3.2:1b)

⚙️ Phase 3: A - Architect (The 3-Layer Build)

  • Layer 1: Architecture (architecture/) - Write SOPs for conversion logic
  • Layer 2: Navigation - Logic to route code chunks to conversion tools
  • Layer 3: Tools (tools/) - Implement Python/Node scripts for actual conversion

✨ Phase 4: S - Stylize (Refinement & UI)

  • Refine output formatting (Prettier, ESLint rules)
  • If applicable, a simple UI for uploading/pasting code

🛰️ Phase 5: T - Trigger (Deployment)

  • Final documentation
  • Maintenance log in gemini.md

Checklists

  • Core conversion logic for Basic Selectors
  • Core conversion logic for Actions (Click, Type)
  • Core conversion logic for Assertions
  • Wait mechanism conversion (Explicit/Implicit -> Playwright auto-waiting)
  • Page Object Model (POM) structure conversion