What problem does it solve? Writing user documentation for a web application is slow and quickly goes stale: someone must manually click through every page, take screenshots, and write step-by-step instructions. This Skill automates that process by browsing the running app via browser automation, screenshotting every screen, and producing a structured, publishable user guide. ## Core Features & Use Cases - Automated App Walkthrough: Uses Chrome MCP, Playwright MCP, or playwright-cli to discover all routes, navigate pages, and capture consistent 1280x720 screenshots with realistic data. - Four Documentation Depths: Choose quick (~10 screenshots, quick-start guide), standard (~30, full user guide), thorough (~80+, all states and flows), or exhaustive (~150+, publishable suite with admin guide, FAQ, and HTML version). - Structured Output: Produces GitHub-flavored Markdown with per-page sections, how-to workflows, keyboard shortcut tables, troubleshooting, workflow diagrams, and an optional self-contained HTML version. - Use Case: After shipping a new client-management dashboard, run the skill at standard depth to produce a complete USER_GUIDE.md with annotated screenshots of every page and CRUD workflow, ready to commit to the repo or hand to clients. ## Quick Start Ask the agent to document your running app at a given URL, for example: generate a standard-depth user guide with screenshots for the app running at http://localhost:3000.
