browserai

The Browserai MCP server acts as a bridge between your AI agents and real-time web data via the Browserai API. It enables MCP clients like Claude Desktop, VS Code, and other compatible agents to perform web searches, navigate websites, and extract data with anti-scraping defenses handled by Browserai. By running this MCP server, you expose a standardized interface that lets your agent issue real-time web actions and receive structured data back for use in prompts, reasoning, or automation workflows.

Once the server is running, configure your MCP client to connect to the Browserai MCP server using the provided command and environment variables. You’ll need your Browserai API token and, optionally, a project name to scope requests. The client will send model-context protocol requests to search the web, browse pages, or extract targeted information, and the MCP server will orchestrate Browserai’s capabilities behind the scenes, returning clean results or structured objects suitable for downstream consumption by the AI agent.

Prerequisites:

• Node.js and npm installed on the host machine. You can download them from https://nodejs.org/
• Access to a Browserai account with an API token (to be used in MC P environment variables)

Installation steps:

1. Install Node.js (if not already installed) by following the installer on nodejs.org.
2. Open a terminal and verify Node.js and npm versions:
   • node -v
   • npm -v
3. Install and run the MCP server using npx (no local install required in most setups):
   • npx @brightdata/browserai-mcp
4. Provide required environment variables when starting the server or configure them in your MCP client as described in the README:
   • API_TOKEN: Your Browserai API token (required)
   • PROJECT_NAME: Optional project name (defaults may apply if omitted)

Optional: If you prefer a local install, you can install the package globally and run it via npm or npx-equivalent commands provided by your environment.

Tips and common considerations:

• Keep your API token secure; avoid hard-coding it in code repositories. Use environment variables or secret management in your deployment environment.
• The PROJECT_NAME environment variable is optional and allows you to target a specific Browserai project; omitting it defaults to a pre-configured project depending on your Browserai setup.
• If you encounter an ENOENT error for npx, ensure Node.js/npm are correctly installed and available in your PATH, or specify the full path to node/npm in your MCP client configuration.
• Timeouts may occur for complex or long-running web actions; adjust your agent’s overall timeout settings accordingly (e.g., 180s or higher depending on tasks).
• Treat scraped web content as potentially untrusted; validate and sanitize data before feeding it into prompts or downstream tooling.