Playwright Stealth
Give your agent Playwright-driven browser automation with optional stealth and headless modes for scraping, UI checks, and flows that hit real pages.
Overview
Playwright Stealth MCP is a MCP server for the Build phase that runs Playwright browser automation with optional stealth and headless configuration over stdio.
What is this MCP server?
- Stdio MCP server playwright-stealth-mcp-server v0.2.2 built on Playwright
- STEALTH_MODE env toggles anti-detection measures; optional custom STEALTH_USER_AGENT and STEALTH_LOCALE
- STEALTH_MASK_LINUX defaults true to present Windows-like UA when masking Linux fingerprints
- HEADLESS defaults true for unattended agent-driven sessions
- npx -y PulseMCP experimental playwright-stealth package
- Server package version 0.2.2, npm identifier playwright-stealth-mcp-server
- STEALTH_MODE default false; HEADLESS default true; STEALTH_MASK_LINUX default true
- Stdio transport with npx -y from PulseMCP experimental playwright-stealth subfolder
Community signal: 70 GitHub stars.
What problem does it solve?
Your agent cannot complete build tasks that require clicking, logging in, or reading dynamic pages when APIs are missing or incomplete.
Who is it for?
Indie builders automating browser workflows during Build who need Playwright plus light anti-detection toggles from MCP env vars.
Skip if: High-volume scraping against strict bot protection, compliance-sensitive access without legal review, or teams that only need static HTTP clients.
What do I get? / Deliverables
After install, your agent can drive a configurable headless or stealth browser session to complete web integration and validation steps from the IDE.
- MCP tools that launch and control Playwright browser sessions
- Configurable stealth and headless behavior via documented environment variables
- Agent-executable web flows for build-time integration and validation
Recommended MCP Servers
Journey fit
Build is canonical because browser automation is used while implementing integrations, validating UIs, and automating web-facing product behavior—not primarily for launch SEO or operate monitoring. Integrations captures MCP-backed control of a real browser for third-party sites, auth flows, and automated interaction during product construction.
How it compares
MCP Playwright browser automation with stealth env flags, not a hosted browser farm or a single-purpose scraper skill.
Common Questions / FAQ
Who is Playwright Stealth MCP for?
Solo developers using agent-first workflows who need Playwright-controlled browsers for integration and UI tasks from Claude Code, Cursor, or Codex.
When should I use Playwright Stealth MCP?
Use it in Build integrations when you must interact with live pages, optional stealth, and headless defaults while implementing or verifying web behavior.
How do I add Playwright Stealth MCP to my agent?
Register the stdio server via npx -y playwright-stealth-mcp-server and set STEALTH_MODE, HEADLESS, and related STEALTH_* variables to match your target sites.