Vibeleaderboard
← Back to Tools
Index / tool

Playwright CLI

https://github.com/microsoft/playwright-cli
Visit playwright.dev
Category
Developer Tools
Rank
No. 169Tools index
Pricing
Open Source
Type
TOOL
Builder
microsoft
GitHub
13.2k stars
Latest release
v0.1.19
Date

About

CLI for common Playwright actions — record browser interactions and have it generate the Playwright code that reproduces them.

What it does

Playwright CLI lets coding agents operate browsers through compact terminal commands. Agents can navigate pages, inspect snapshots, find elements, interact with controls, manage tabs and storage, mock network requests, capture screenshots or PDFs, and inspect console and request activity. Browser state can survive between commands within named sessions.

Why it's ranked here

Its unusually broad command surface makes it a practical browser operator for coding agents, not merely a thin launcher. Snapshot depth controls, named sessions, persistent profiles, tracing, network routing, and a live monitoring dashboard cover both routine interaction and deeper debugging. The project also states a clear design goal: keep page data out of model context unless requested.

What's good

The session model is especially useful. Cookies and storage remain available across separate terminal invocations, while optional disk persistence survives browser restarts. Multiple named browsers can serve different projects, and the dashboard provides live previews plus manual mouse and keyboard takeover. Agents can also narrow snapshots by depth or search their text, reducing unnecessary page data.

Tradeoffs

Node.js 18 or newer is required, and installation uses a global npm package. Browser profiles live only in memory unless disk persistence is explicitly enabled. The CLI favors concise, purpose-built commands over continuous page introspection. The repository itself positions MCP as the better adjacent option for persistent agent loops, rich structural inspection, exploratory automation, and self-healing tests.

How to use it well

Use it with coding agents that need browser checks alongside implementation work, especially when context space matters. Install the supplied skills, assign a named session per project, keep the default temporary profile for disposable tests, and enable persistence only when login or storage must survive restarts. Use the dashboard when humans need to observe or take control. Choose Playwright MCP instead when continuous browser context and rich page reasoning matter more than compact commands.

Technical notes+

The README documents a global npm install for @playwright/cli, a playwright-cli binary, and optional local SKILL installation. The command surface covers navigation, element references obtained from snapshots, keyboard and mouse input, tabs, cookies, local and session storage, request routing, console and network inspection, screenshots, PDFs, tracing, video, and arbitrary Playwright snippets. Sessions retain profiles in memory by default, support disk persistence, and can be selected through a flag or PLAYWRIGHT_CLI_SESSION. LICENSE contains the Apache License 2.0 text and identifies Microsoft Corporation.

Observed

License
Apache License 2.0
Installation
Global npm package: @playwright/cli
Interface
Command-line interface for Playwright browser automation
Runtime requirement
Node.js 18 or newer
Agent integrations
Optional locally installed skills for Claude Code, GitHub Copilot, and other coding agents
Browser operation
Headless by default, with an optional headed mode
Session storage
In-memory profiles by default, with optional disk persistence
Monitoring interface
Visual dashboard for viewing and controlling running browser sessions

Read from LICENSE.

What it can do

  • Record browser interactions and generate Playwright code

    Live browser session with user interactions → Playwright test code that reproduces the interactions

  • Generate test scripts from recorded actions

    Recorded browser interaction session → Executable Playwright test script

  • Capture user interface elements during recording

    Browser elements clicked/interacted with → Element selectors and interaction code

  • Convert manual testing workflows to automated tests

    Manual browser testing steps → Automated Playwright test suite

  • Execute common Playwright operations via command line

    CLI commands and parameters → Playwright actions executed in browser

Tags

playwrightclimicrosofttestingautomation

Tech Stack

Node.js

Comments (0)

No comments yet

Editorially curated, with community endorsements as a secondary signal. Corrections welcome.