# Changelog

All notable changes to Playwright Pilot AI will be documented in this file.

The format follows [Keep a Changelog](https://keepachangelog.com/en/1.0.0/).
This project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0.html).

---

## [0.1.0] — 2026-06-04

### Added
- `pwpilot_preview_generation` — shows a full generation plan before any browser or AI action
- `pwpilot_launch_browser` — opens a visible Chromium browser at the target URL
- `pwpilot_analyze_page` — extracts all interactive elements from the live DOM (zero AI tokens)
- `pwpilot_generate_test` — runs local skills to build the POM, then asks Claude to write only the spec (~600 tokens)
- `pwpilot_save_test` — writes POM, spec, StepTracker helper, intent JSON, and YAML definition to disk
- `pwpilot_video_to_test` — converts a screen recording or user-supplied steps into a Playwright test
- `pwpilot_run_yaml` — executes a test directly from its YAML definition (no spec file needed)
- `pwpilot_execute_test` — runs a saved test intent with self-healing
- `pwpilot_heal_test` — repairs broken locators using stored ElementContext metadata
- Auto-scaffolding of production-standard JavaScript (Node.js CommonJS) Playwright projects
- YAML portable test definitions with primary + fallback locators for self-healing
- StepTracker runtime helper (JavaScript and TypeScript versions)
- `playgen` and `playwright-pilot-mcp` CLI binaries
- CLAUDE.md routing instructions for Claude Desktop and Cursor