Index live · v1.7.0 · OCT 6 2026

Playwright

Official
byMicrosoft

Microsoft's official Playwright MCP server enables AI agents to automate browser interactions using structured accessibility snapshots rather than screenshots. This approach is more reliable and faster than vision-based automation. Perfect for testing, scraping, and web automation workflows.

stdioTypeScript
27.3Kstars

Configuration

{
  "mcpServers": {
    "playwright": {
      "command": "npx",
      "args": [
        "@playwright/mcp@latest"
      ]
    }
  }
}

Run in a terminal. Copied verbatim from the upstream README.

AgenticSkills Audit

Automated framework checks. Deep code review tracked separately. Read the methodology →

6/7
6 of 7 automated checks passed
Audited Aug 7, 2026
SECURITY.md published
Mitigates: Coordinated disclosure path
Evidence
Commit in last 90 days
Mitigates: Maintainer abandonmentLast push 1 day(s) ago
≥2 active contributors
Mitigates: Bus-factor of one10 contributors with ≥2 commits
Evidence
CI pipeline configured
Mitigates: Code health regression
Evidence
Dependency lockfile committed
Mitigates: Dependency confusionpackage-lock.json
Evidence
Software Bill of Materials (SBOM)
Mitigates: Audit trailNo SBOM file present
License declared
Mitigates: Legal ambiguityApache-2.0

Deep framework checks (OAuth 2.1 / PKCE implementation, input validation, sandboxing) require human source review and are tracked separately. This scorecard covers programmatically verifiable signals only.

Quick Stats

Trust LevelOfficial
Stars27,300
Transportstdio
API KeyNot required
LanguageTypeScript
CategoryDeveloper Tools
Sourcemicrosoft/playwright-mcp
Websiteplaywright.dev

Author

M

Microsoft

Official

Tags

browsertestingautomationweb-scraping

Get new MCP servers in your inbox. What we added, what failed the security audit, and why. See a past issue.