terminal_skill

Overview

This skill automates browser interactions from the command line using Playwright. It provides streamlined commands to navigate pages, interact with elements, take screenshots, run multi-tab workflows, and capture traces for debugging. Use it to script end-to-end tasks without writing full test suites.

How this skill works

playwright-cli controls Chromium, Firefox, and WebKit through a simple CLI command set. You open URLs, reference page elements from snapshots, and execute actions like click, fill, upload, and keyboard/mouse events. Sessions, tracing, and DevTools commands let you persist state, record diagnostics, and inspect network/console output.

When to use it

• Quickly prototype or reproduce browser flows from the terminal
• Automate form filling, uploads, and data extraction without code-heavy scripts
• Capture screenshots or PDFs of pages and specific elements
• Debug UI issues with tracing, network inspection, and console logs
• Run multi-tab scenarios and preserve browser sessions for repeatable workflows

Best practices

• Start by running snapshot to reference stable element ids before interacting
• Use sessions to store cookies and authentication state for repeated runs
• Prefer explicit waits or re-snapshot after significant page changes to avoid flakiness
• Capture traces and network logs when investigating intermittent failures
• Limit viewport/resolution with resize to reproduce layout-specific bugs

Example use cases

• End-to-end smoke test: open an app, fill login form, click submit, snapshot the result
• Data extraction: navigate pages, eval element text, and save screenshots or PDFs
• Form automation: upload files, fill inputs, select options, and verify results with snapshot
• Multi-tab flow: open multiple pages, switch between tabs, and perform coordinated actions
• Debugging: start tracing, reproduce the issue, stop tracing and inspect network/console logs