Frequently Asked Questions (FAQ) about browser-use MCP Server

Q: What is the browser-use MCP Server? A: The browser-use MCP Server is an AI-driven browser automation server that implements the Model Context Protocol (MCP) for natural language browser control and web research. It allows AI models to interact with web browsers using natural language commands.

Q: What is MCP? A: MCP stands for Model Context Protocol. It is an open protocol that standardizes how applications provide context to Large Language Models (LLMs).

Q: What are the key features of the browser-use MCP Server? A: Key features include MCP integration, browser automation, visual understanding (screenshot analysis), state persistence, multi-LLM support, a deep research tool, environment variable configuration, and CDP connection.

Q: Which LLM providers are supported? A: The server supports OpenAI, Anthropic, Azure, DeepSeek, Google, Mistral, Ollama, OpenRouter, Alibaba, Moonshot, and Unbound AI.

Q: What is the run_browser_agent tool? A: The run_browser_agent tool executes a browser automation task based on natural language instructions and waits for it to complete.

Q: What is the run_deep_search tool? A: The run_deep_search tool performs in-depth web research on a topic, generates a report, and waits for completion.

Q: How do I configure the server? A: The server is configured using environment variables. You can set these in your system or place them in a .env file in the project root.

Q: What is CDP and how do I connect to my own browser? A: CDP stands for Chrome DevTools Protocol. You can connect to your own Chrome/Chromium browser by launching it with remote debugging enabled and setting the MCP_USE_OWN_BROWSER and CHROME_CDP environment variables.

Q: What if I get API errors? A: Double-check that the correct API key environment variable is set for your chosen MCP_MODEL_PROVIDER, or that MCP_API_KEY is set. Verify keys and endpoints (AZURE_OPENAI_ENDPOINT is required for Azure).

Q: How do I troubleshoot browser conflicts? A: If not using CHROME_CDP (MCP_USE_OWN_BROWSER=false), ensure no other conflicting Chrome instances are running with the same user data directory if CHROME_USER_DATA is specified.

Q: How does this integrate with UBOS? A: The browser-use MCP Server integrates with UBOS by allowing you to create AI Agents within the UBOS platform that utilize the server to interact with web browsers. You can then orchestrate these agents within UBOS workflows.

Q: What are the prerequisites for running the server? A: You need Python 3.11 or higher, uv (fast Python package installer), and Chrome/Chromium browser installed with Playwright browsers installed.

Q: Where can I find the log file? A: The log file path is specified by the LOG_FILE environment variable (default: mcp_server_browser_use.log).

Q: What is the difference between BROWSER_HEADLESS and MCP_HEADLESS? A: BROWSER_HEADLESS primarily affects run_deep_search, while MCP_HEADLESS specifically affects the run_browser_agent tool. If MCP_USE_OWN_BROWSER=true, both are ignored.