MCP Server
Enables interaction with the BrowserBase SDK for Python, allowing agents to programmatically control and automate web browsers, as specified in the OpenAPI URL.
Click on "Deploy Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type
@followed by the MCP server name and your instructions, e.g., "@MCP Serverstart a new browser session and navigate to example.com"
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
MCP Server
This project is an MCP (Multi-Agent Conversation Protocol) Server for the given OpenAPI URL - https://raw.githubusercontent.com/browserbase/sdk-python/refs/heads/main/openapi.v1.yaml, auto-generated using AG2's MCP builder.
Prerequisites
Python 3.9+
pip and uv
Related MCP server: Playwright Browserbase MCP Server
Installation
Clone the repository:
git clone <repository-url> cd mcp-serverInstall dependencies: The .devcontainer/setup.sh script handles installing dependencies using
pip install -e ".[dev]". If you are not using the dev container, you can run this command manually.pip install -e ".[dev]"Alternatively, you can use
uv:uv pip install --editable ".[dev]"
Development
This project uses ruff for linting and formatting, mypy for static type checking, and pytest for testing.
Linting and Formatting
To check for linting issues:
ruff checkTo format the code:
ruff formatThese commands are also available via the scripts/lint.sh script.
Static Analysis
To run static analysis (mypy, bandit, semgrep):
./scripts/static-analysis.shThis script is also configured as a pre-commit hook in .pre-commit-config.yaml.
Running Tests
To run tests with coverage:
./scripts/test.shThis will run pytest and generate a coverage report. For a combined report and cleanup, you can use:
./scripts/test-cov.shPre-commit Hooks
This project uses pre-commit hooks defined in .pre-commit-config.yaml. To install the hooks:
pre-commit installThe hooks will run automatically before each commit.
Running the Server
The MCP server can be started using the mcp_server/main.py script. It supports different transport modes (e.g., stdio, sse, streamable-http).
To start the server (e.g., in stdio mode):
python mcp_server/main.py stdioThe server can be configured using environment variables:
CONFIG_PATH: Path to a JSON configuration file (e.g., mcp_server/mcp_config.json).CONFIG: A JSON string containing the configuration.SECURITY: Environment variables for security parameters (e.g., API keys).
Refer to the if __name__ == "__main__": block in mcp_server/main.py for details on how these are loaded.
The tests/test_mcp_server.py file demonstrates how to start and interact with the server programmatically for testing.
Building and Publishing
This project uses Hatch for building and publishing. To build the project:
hatch buildTo publish the project:
hatch publishThese commands are also available via the scripts/publish.sh script.
Related MCP Connectors
Hyperbrowser MCP — wraps the Hyperbrowser AI-agent browsing API
Provides cloud browser automation capabilities using Stagehand and Browserbase, enabling LLMs to i…
MCP server exposing the Backtest360 engine API as tools for AI agents.
MCP server for uagents documentation, generated by doc2mcp.
Related MCP Servers
- FlicenseNot gradedqualityDmaintenanceA Multi-Agent Conversation Protocol server that provides a standardized interface for agent interactions with a FastAPI service, auto-generated using AG2's MCP builder.-
- AlicenseAqualityDmaintenanceAn MCP server that enables LLM applications to control web browsers via Browserbase, supporting features like web navigation, screenshots, cookie management, and persistent contexts.172,298 npmApache 2.0
- AlicenseCqualityBmaintenanceA professional browser automation server that enables MCP clients to perform structured web navigation, element interaction, and data extraction using the DrissionPage framework. It features 14 deterministic tools optimized for LLMs to automate web workflows efficiently without relying on vision-based models.6984 PyPI488Apache 2.0
- AlicenseNot gradedqualityDmaintenanceA Selenium-based MCP server that exposes browser automation tools for navigation, interaction, form filling, and assertions, enabling AI agents to control web browsers through natural language.GPL 3.0