OpenViSTAI · Agents · Blockchain · Crypto
Menu
MCP & Tooling

Browser Use MCP Server

Browser Use MCP Server is a Python server that exposes browser-use browser automation to AI agents through the Model Context Protocol. It supports SSE and stdio transports and lists asynchronous tasks and VNC streaming among its features.

Independent OpenViST project profile, based on linked sources.

Browser Use MCP Server — OpenViST project profile in MCP & Tooling.
OpenViST project illustration. See the linked sources for product details.
Understand the project

Practical overview

When it fits

Use it when an MCP-compatible agent needs to interact with a browser, such as opening a page and retrieving an item from it, and you can run and configure the supporting server locally or in a self-hosted environment.

Who it is for

  • Developers connecting browser automation to MCP-compatible AI clients.
  • Cursor, Windsurf, or Claude Desktop users configuring an agent to interact with web pages.
  • Teams building agent applications that need browser interaction as a component.

Documented uses

  • Give an AI agent a way to navigate web pages and return requested information.
  • Connect an MCP client to a browser-control server using SSE or stdio.
  • Observe browser automation through the documented VNC streaming option.
Before you choose

Requirements & limits

The repository is public and identifies the project as MIT-licensed. The documented setup uses local server commands and configuration; the README specifies an OpenAI API key in its environment and client examples.

  • Choose SSE or stdio based on the client connection method; stdio requires mcp-proxy.
  • The setup installs Chromium through Playwright and supports an optional custom Chrome path.
  • VNC streaming and asynchronous tasks are listed as features; Docker instructions describe exposing server and VNC ports.

Check current requirements in the official documentation. Source-based guidance does not mean OpenViST has run or independently tested this project.

Source-based practical guide

Install the server prerequisites and browser

Prepare the documented Python and browser automation dependencies before configuring a client.

Before you start

  • uv
  • Playwright
  • mcp-proxy if using stdio mode

Steps

  1. Install uv and install mcp-proxy with uv if you plan to use stdio.

  2. From the project directory, run the command below.

    Documented command
    uv sync
  3. Install the Playwright package with the command below.

    Documented command
    uv pip install playwright
  4. Install Chromium and its dependencies with the command below.

    Documented command
    uv run playwright install --with-deps --no-shell chromium

Expected outcome

The listed prerequisites and Chromium browser are installed for proceeding to server startup and client configuration.

Guide limitations

  • This covers dependency setup only; it does not verify a successful server launch.
  • mcp-proxy is required for stdio mode, not listed as a prerequisite for SSE mode.
Source-based practical guide

Configure an MCP client for SSE mode

Point an MCP client at the server's documented SSE endpoint.

Before you start

  • The server's documented dependencies installed
  • An MCP client that accepts the shown server configuration

Steps

  1. Start the server from source with the command below.

    Documented command
    uv run server --port 8000
  2. Add an MCP server entry with the URL http://localhost:8000/sse.

  3. For Cursor, save the configuration at ./.cursor/mcp.json.

Expected outcome

The client configuration targets the SSE endpoint on the local server.

Guide limitations

  • The configuration example assumes the server is reachable at localhost on port 8000.
Evidence and transparency

Sources & review

Source record details
Source availability
Active
Sources checked
12 October 2026
Original website title
kontext-security/browser-use-mcp-server · Official homepage title · retained in its source language