MCP server for Playwright

MCP server for Playwright is an application that provides AI assistants with browser automation and web page interaction capabilities via the standard Model Context Protocol (MCP). The server uses structured accessibility snapshots of pages and does not require vision-capable models. It is deployed on a virtual machine in Yandex Cloud.

Key Features and Benefits

• Yandex AI Studio Integration. The MCP server integrates with AI Studio MCP Servers — connect it to your AI agents in just a few clicks.
• Browser Automation. Navigate pages, click elements, type text, fill forms, and perform other actions in Chromium through an AI assistant.
• Accessibility Page Snapshots. Work with a structured representation of the page without the need for image recognition.
• Headless Mode. The browser runs without a GUI for efficient operation on servers.
• Content Extraction. Retrieve text, links, forms, and other elements of web pages.
• Screenshots. Capture screenshots of web pages.
• Bearer Token Authorization. All requests to the MCP server are protected with a Bearer token stored in Yandex Lockbox.
• Standard MCP Protocol. Uses the open Model Context Protocol for AI model interaction.
• Streamable HTTP Transport. The MCP server is accessible on port 8000 using Streamable HTTP.
• Minimal Resources. The application runs on a virtual machine with minimal configuration (2 vCPU, 2 GB RAM).
• Docker Containerization. The server runs in Docker containers (MCP + nginx auth proxy).

1. Create a secret in Yandex Lockbox to store the Bearer token for authorization. The secret must contain a bearer_token key with an arbitrary token value.
2. Create a service account for deploying the application and assign it the roles compute.editor, iam.serviceAccounts.admin, and vpc.admin on the deployment folder.
3. Assign the installer service account the lockbox.admin role separately on the Bearer token secret. This role is required to read and modify access permissions on the selected secrets; no folder-level Lockbox role is required. This also applies to secrets in another folder. During installation, the application creates a dedicated VM service account and grants it lockbox.payloadViewer only on the selected secrets.
4. In the management console select the Cloud Apps service.
5. In the left panel, select Application Store.
6. Select MCP Playwright and click the Use button.
7. Specify:
   • Application name.
   • (Optional) Application description.
   • The service account created earlier.
   • The cloud subnet where the virtual machine will be deployed.
   • Lockbox secret ID containing the Bearer token for authorization.
   • External access to the MCP server (optional; disabled by default, enable for connections through the public IP address).
   • SSH access. If you select Enabled, specify the required Username and SSH public key.
8. Click the Install button and wait for the application to install.
9. After installation, choose the appropriate URL in Application data: MCP server URL (internal) for clients with network access to the VM’s internal IP address, or MCP server URL (external) for the supported public-access scenario. In either case, configure the client with the Authorization: Bearer <your-token> header.

Connecting the MCP server to Yandex AI Studio

1. In the management console open AI Studio and go to Agent Core → MCP servers.
2. Click Create MCP server and select the required MCP server under MCP servers from Yandex Cloud Marketplace.
3. Specify the server name and select an existing application instance, or create a new one by following the deployment instructions. When deploying a new server, the Save button becomes available only after the MCP server has been fully deployed and started.
4. After you select the application, the MCP server’s Network and Service account settings are filled in automatically.
5. Click Create. After the connection has been verified successfully, select the required tools and click Save. The MCP server will be registered in AI Studio and available for use by AI agents.

• Automating routine browser actions via an AI assistant: filling forms, navigating sites, collecting data.
• Extracting information from web pages using accessibility snapshots without image recognition.
• End-to-end testing of web applications with scenarios controlled via natural language.
• Capturing screenshots of pages for reports and archiving.
• Building AI agents capable of interacting with modern web applications.

Yandex Cloud technical support is available 24/7. The types of requests you can submit and the appropriate response time depend on your pricing plan. You can switch to the paid support plan in the management console. You can learn more about the technical support terms here.