What problem does it solve? Iterating on a web game without automated feedback leads to broken controls, visual regressions, and undetected console errors. This Skill enforces a disciplined implement-act-pause-observe loop so every change to an HTML/JS game is verified through real browser interaction before moving on. ## Core Features & Use Cases - Playwright Action Loop: Runs scripted keyboard and mouse input bursts against the game via a Node.js client, capturing screenshots, text state, and console errors after each iteration. - Deterministic Time Stepping: Uses a window.advanceTime(ms) hook and virtual-time shim so frames advance predictably during automated tests instead of relying on wall-clock timing. - Text-Based State Inspection: Reads a window.render_game_to_text JSON payload to verify game state (positions, score, entities) matches what screenshots show. - Progress Handoff: Maintains a progress.md file with the original prompt, TODOs, and notes so another agent can continue the work seamlessly. - Use Case: Ask the agent to add a jump mechanic to a platformer; it implements the change, runs the Playwright client with arrow-key and spacebar payloads, inspects the screenshots and state JSON, fixes any console errors, and repeats until the mechanic works end-to-end. ## Quick Start Ask the agent to build a small canvas game in this workspace and validate each change by running the Playwright test client with action payloads, reviewing screenshots and console errors after every iteration.
