Agent connectors quickstarts

These tutorials get you started using Airbyte's agent connectors.

📄️ Open source Python SDK

In this tutorial, you'll create a new Python project with uv, add a Pydantic AI agent, equip it to use one of Airbyte's agent connectors, and use natural language to explore your data. This tutorial uses GitHub, but if you don't have a GitHub account, you can use one of Airbyte's other agent connectors and perform different operations.

📄️ Open source Connector MCP

In this tutorial, you'll install and run Airbyte's connector MCP server locally, connect the MCP server to Claude Code or your preferred agent, and learn to use natural language to explore your data. This tutorial uses Stripe, but if you don't have a Stripe account, you can use one of Airbyte's other agent connectors.

📄️ Hosted connectors

When you run connector operations with the Python SDK, you store API credentials locally and provide them directly to the API through agent connectors. This approach, while viable at first, reaches limits quickly. You may find yourself dealing with large numbers of customers who have their own environments and credentials. You may not want to manage these credentials and or store them at all.

📄️ Implement your own OAuth flow

If you provide an environment where your users can create connectors, they need to supply their credentials so agents can access their data. Airbyte provides a standard Embedded widget for this purpose. However, you might prefer to create a fully customized OAuth flow with your own branding and UX. In this case, you can implement your own OAuth flow.