Meet Goose: The open source AI agent built for developers

From debugging to automation, with 1,700+ MCP servers ready to plug in.

AI coding assistants are evolving fast, but few go beyond code generation. In her lightning talk at All Things Open, Angie Jones from Block shares Goose, an open source AI agent designed to run on your local machine and interact with your development environment like a true assistant.

Read more: AI agents are here—are your skills ready?

Angie begins by highlighting the gap between chat-based assistants and practical developer workflows. While tools like ChatGPT can suggest fixes, they lack direct access to your codebase, leaving developers to manually test and retry. Goose changes that by connecting to your IDE and running commands on your behalf. Instead of telling you how to fix an error, it just fixes it.

She then explains how Goose works under the hood. The agent connects to the large language model (LLM) of your choice, whether commercial models or local open source options, and augments them with tools. These tools allow the agent to interact with files, run tests, and execute actions within your environment. Through the Model Context Protocol (MCP), Goose can extend its capabilities with thousands of available extensions, from GitHub automation to Google Drive integration.

Angie also shares how Goose scales beyond debugging. With extensions, it can generate code from Figma designs, manage GitHub repos, or even plan around a parade route with Google Maps. Goose’s flexibility and open architecture encourage developers to contribute new tools, making it a rapidly growing ecosystem. With more than 1,700 MCP servers available and support for multiple programming languages, the possibilities are wide open.

Key takeaways

• AI agents like Goose move beyond suggestion, handling real tasks directly in your development environment.
• The Model Context Protocol makes Goose extensible, enabling developers to connect thousands of tools and APIs.
• As an open source project, Goose invites community contributions, offering a path to innovation no closed system can match.

Goose shows what happens when AI agents run locally, connect seamlessly to your tools, and are supported by a strong open source community. Angie’s talk highlights how developers can not only use Goose today but also help shape its future.

More from We Love Open Source

• Deep dive into the Model Context Protocol
• Discover Goose: Automate your developer setup with this AI agent
• Why AI agents are the future of web navigation
• What if your AI agent could actually help?
• AI agents are here—are your skills ready?