Skip to content
← Tools / Skills
Skill · .md packageMicrosoft/10 trustVerified

Playwright CLI Skill

Microsoft's official token-efficient Playwright CLI for coding agents, with a bundled SKILL.md.

Visit official site →

About

microsoft/playwright-cli is Microsoft's official agent-first browser automation tool (npm @playwright/cli), and the repo ships an official skill at skills/playwright-cli/SKILL.md (plus ~10 reference files). The README explicitly positions CLI+SKILLs as the better fit for coding agents than MCP (more token-efficient: no large tool schemas or verbose accessibility trees in context).

What it can do

  • One-shot browser commands (open/goto, click/fill/type/press/hover/select/check via accessibility refs)
  • Snapshot-based element discovery; screenshots, PDFs, tracing, video recording
  • Console/network/devtools inspection; request mocking (route/unroute)
  • Storage-state save/load for auth sessions
  • Multi-tab sessions and headed/headless operation
  • Generating and debugging Playwright test code

Setup

  1. npm install -g @playwright/cli@latest
  2. playwright-cli install --skills (installs the skill into Claude Code / Copilot / other detected agents), or npx skills add microsoft/playwright-cli@playwright-cli
  3. Verify: playwright-cli --help

Prerequisites

Node.js 18+; a Chromium/Firefox/WebKit browser (installable via the CLI).

Use cases

testingautomationvision

Works with

Claude CodeCopilotAny agent (.md)

Trust score — 9/10

/10 trustVerified

Official Microsoft, Apache 2.0, actively maintained (25 releases, ~14K stars), purpose-built for agents with thorough docs. Caveat is functional: it drives real browsers that can act on live sites — keep the agent within authorized URLs, accounts, and mutation scope.

★ 14,053 GitHub stars

How the trust score is built

Full methodology →
Adoption · stars/downloads, log-scaled vs peers
7.8/10
Maintenance · recency, releases, issue load
9.5/10
Openness · license + docs
10.0/10
Manual review · sanity read + red-flag check
9.0/10
  • · 14,053 GitHub stars (log-scaled vs catalog peers)
  • · last push 14d ago; 25 releases/12mo; 0 open issues
  • · manual review: drives real browsers — functional caution, not a defect

Sources

Every fact above was checked against these sources. Where a claim comes from the provider and couldn't be independently verified, the text says so.