playwright

Automate real browsers from the terminal using playwright-cli commands.

Updated Jun 2, 2026
One-click install
npx skills add https://github.com/IagoPrandi/ai-dealer --skill playwright-iagoprandi
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: playwright
Source: https://github.com/IagoPrandi/ai-dealer/tree/main/.claude/skills/playwright
Command: npx skills add https://github.com/IagoPrandi/ai-dealer --skill playwright-iagoprandi

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires @playwright/mcp, and includes scripts (resource) and references (resource) components.

What problem does it solve? Manual browser testing and UI verification are slow and error-prone. This Skill lets you drive a real browser from the terminal to navigate pages, fill forms, take snapshots and screenshots, extract data, and debug UI flows without writing test files. ## Core Features & Use Cases - CLI-first browser automation: Open pages, snapshot the DOM to get stable element refs, click, type, fill forms, and capture screenshots or PDFs via playwright-cli or the bundled wrapper script. - Container-based execution: Run Playwright inside the official Linux container image for consistent browser checks on Windows workspaces, including attaching to Docker Compose networks. - Debugging and inspection: Capture console output, network activity, and traces around suspicious flows, with named sessions to isolate work. - Use Case: Verify a local web app by opening http://localhost:3000, snapshotting the page, filling a login form, clicking submit, and saving a screenshot under output/playwright/. ## Quick Start Ask the AI to open a URL with the Playwright wrapper, snapshot the page, and take a screenshot of the result.

Frequently Asked Questions about playwright

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I automate a browser from the command line with Playwright?▼

Use playwright-cli through the wrapper script, which runs `npx --package @playwright/mcp playwright-cli` without a global install. Open a page, run `snapshot` to get element refs, then use commands like `click`, `fill`, and `screenshot`.

How to run Playwright in Docker on a Windows workspace?▼

Run the official mcr.microsoft.com/playwright container image with the repo mounted at /work and execute pnpm test:e2e or playwright screenshot inside it. This avoids depending on host-installed browsers or host Chrome/Edge.

Does playwright-cli require a global installation?▼

No, the bundled wrapper script uses npx to fetch @playwright/mcp on demand, so only Node.js/npm with npx is required. A global install of playwright-cli is optional convenience.

Why do Playwright element refs fail after clicking?▼

Refs go stale after navigation, modal changes, or significant DOM updates. Run `snapshot` again to get fresh element refs and retry the interaction with the new identifiers.

What to do when agent-browser is not installed on PATH?▼

Run it through `npx -y agent-browser` instead of a global command. If no browser is found, use `agent-browser doctor` and `agent-browser install`, or pass an explicit `--executable-path` from the Playwright browser cache.