[ MCP SERVER SUMMARY ]
Playwright MCP server summary
Microsoft's Playwright MCP server gives AI assistants browser automation through accessibility snapshots — navigate pages, click, fill forms, and run deterministic web workflows without screenshot-based vision models. Bifrost centralizes the server process, scopes which automation tools each team can call, and audits browser actions alongside LLM traffic.
| Property | Details |
|---|---|
| Company | Microsoft |
| Connection type | STDIO |
| Authentication | None (local browser automation) |
| Package or endpoint | @playwright/mcp@latest |
| Capabilities | Browser navigation and interaction, Accessibility-tree snapshots, Form filling and click automation |
| Official documentation | Playwright MCP documentation |
Before you start
- Node.js 18+
- An MCP-compatible client pointed at Bifrost
What you can do with Playwright
- Let coding agents explore and verify web UIs without installing Playwright MCP on every laptop
- Restrict destructive browser tools (downloads, uploads) per virtual key
- Audit agent-driven browser sessions next to model requests for compliance reviews
[ HOW TO CONNECT ]
Connect Playwright to Bifrost in three steps
Install Bifrost
Start the gateway locally or deploy in your VPC.
npx -y @maximhq/bifrost
Add Playwright in MCP Gateway
Open MCP Gateway in the Bifrost dashboard, click New MCP Server, and choose STDIO. Or register via the API below.
• Use STDIO with command npx and args ["@playwright/mcp@latest"].
• No API key is required; Bifrost spawns the Playwright MCP process locally.
Point clients at Bifrost /mcp
Configure Claude Desktop, Cursor, or Claude Code to use your Bifrost endpoint with a virtual key instead of connecting directly to Playwright.
{
"mcpServers": {
"bifrost": {
"url": "http://localhost:8080/mcp",
"headers": {
"Authorization": "Bearer vk_your_virtual_key"
}
}
}
}[ EXAMPLE CONFIG ]
Register Playwright via API
Add this MCP client with a POST to /api/mcp/client. See Connecting to MCP Servers for full options.
curl -X POST http://localhost:8080/api/mcp/client \
-H "Content-Type: application/json" \
-d '{
"name": "playwright",
"connection_type": "stdio",
"stdio_config": {
"command": "npx",
"args": [
"@playwright/mcp@latest"
]
},
"tools_to_execute": [
"*"
]
}'[ WHY BIFROST ]
Run Playwright through one governed gateway
Access control
Virtual keys, RBAC, and tool allowlists per team or agent.
Cost governance
Attribute MCP tool spend alongside LLM usage in one view.
Observability
Log every tool call with latency, server, and virtual key metadata.
Credential isolation
Upstream secrets stay in the gateway — not on every laptop.
[ FAQ ]
Common questions
What does the Playwright MCP server do?
Playwright exposes browser navigation and interaction, accessibility-tree snapshots, form filling and click automation to MCP-compatible AI clients. Bifrost connects to it as an upstream MCP client and exposes approved tools through one governed /mcp endpoint.
How does Playwright authenticate when used with Bifrost?
Playwright uses None (local browser automation). Store credentials in Bifrost's environment or credential configuration so they stay at the gateway instead of being copied into every AI client.
Can Claude Desktop, Cursor, or Claude Code use Playwright through Bifrost?
Yes. After registering Playwright in Bifrost, point any HTTP MCP-compatible client at the Bifrost /mcp endpoint and authenticate with a virtual key. The client sees only the tools allowed for that key.
[ RELATED SERVERS ]
More MCP integrations
Kernel
Provide AI agents with browser infrastructure and cloud browser automation through Kernel’s MCP server.
Sentry
Use Sentry’s MCP server to investigate errors, issues, traces, and application health from AI coding assistants.
Context7
Provide AI coding assistants with current, version-specific library documentation and code examples through Context7.