Playwright (Automation + MCP + Scraper)

# Playwright (Automation + MCP + Scraper) Automate real browser workflows, capture screenshots, and perform reliable web scraping using Playwright and MCP-driven tools. The Playwright (Automation + MCP + Scraper) skill enables AI agents to interact with the web using a real browser engine. Unlike static fetch methods, it handles JavaScript-rendered pages , multi-step forms, and complex UI interactions. It is designed for tasks that require a high degree of reliability, such as end-to-end testing, visual evidence capture, and structured data extraction from dynamic websites. ## How It Works - **Model Context Protocol (MCP)**: . The agent uses specific tools like - **actionability checks**: and resilient locators to ensure that interactions only occur when the UI is ready, significantly reducing automation flakiness. ## Key Features - **MCP-First Browser Control**: Seamlessly navigate, click, fill forms, and select options via standardized agent tools. - **Visual Artifacts**: Capture full-page screenshots, generate PDFs, and record execution traces for debugging. - **Resilient Locators**: Uses web-first assertions and user-visible signals (roles, labels, text) instead of brittle CSS paths. - **Developer Workflow Support**: Built-in support for codegen to record flows and a full test harness for E2E suites. ## Requirements - **Web Connectivity**: Required to access and automate target websites and download browser binaries from NPM. ## Use Cases - **E2E Testing**: Automate complex user journeys and verify UI outcomes across different browser environments. - **Dynamic Web Scraping**: Extract structured data from modern SPAs and pages that require user interaction or authentication. - **UI Bug Reproduction**: Use headed runs and traces to capture and analyze visual regressions or functional failures. ## Installation Install via: `npx clawhub@latest install playwright`

Instalación

Habilidades relacionadas