Jack Dorsey’s Block Unveils Goose, an Open-Source AI Agent

Jack Dorsey’s fintech company Block has introduced Goose, an open-source AI agent framework designed to help developers automate tasks across multiple platforms. Built with transparency and flexibility in mind, Goose allows users to integrate their preferred large language model (LLM), run tasks locally, and connect to various developer tools via extensions.

According to Block’s announcement, Goose can function as an autonomous AI assistant, capable of debugging, code migration, performance benchmarking, and API scaffolding. While its initial focus is on engineering tasks, Block hints that Goose could expand into creative and non-engineering applications in the future.

How Goose Works

Goose is designed to be a versatile, self-sufficient AI agent with the following key features:

• Runs Locally: Unlike cloud-based AI tools, Goose operates directly on the user's machine, providing better control over security and performance.

• Extensible Through Integrations: Developers can connect Goose to GitHub, Google Drive, JetBrains IDEs, and other popular platforms via the Model Context Protocol (MCP).

• Customizable LLM Selection: Users can choose their preferred AI model, with Block recommending Anthropic’s Claude 3.5 Sonnet and OpenAI’s o1 model for optimal performance.

• Dual Interface: Goose is available as both a desktop application and a command-line interface (CLI), ensuring seamless usability across different workflows.

Practical Use Cases

Goose is designed to handle a wide range of tasks, from automating software development workflows to enhancing code quality and performance. Some key use cases include:

• Code Migration: Convert projects between frameworks and languages, such as Ember to React, Ruby to Kotlin, and Prefect-1 to Prefect-2.

• Onboarding to New Codebases: Quickly understand and navigate projects written in unfamiliar programming languages.

• Refactoring & Code Structure Improvements: Transition a codebase from field-based injection to constructor-based injection in a dependency injection framework.

• Performance Optimization: Conduct performance benchmarks for build commands using automation tools.

• Testing & Code Coverage: Generate unit tests for specific features and increase code coverage above a set threshold.

• API & Infrastructure Management: Scaffold APIs for data retention, create Datadog monitors, and manage feature flags (adding or removing them as needed).

• Debugging & Refactoring: Identify and fix issues in unfamiliar codebases, optimizing structure and performance for better maintainability.

While Goose is initially geared toward software engineers, Block envisions broader applications in creative fields, such as AI-assisted music generation.

Why Goose Matters

Block, best known for Square, Cash App, and music streaming service Tidal, isn’t traditionally associated with AI development. However, Jack Dorsey has long been an advocate for open-source software, and Goose represents a step toward making AI-powered automation accessible to more developers and industries.

Goose also arrives amid a growing movement toward open-source AI, following similar efforts from companies like Hugging Face and Stability AI. Some experts believe Goose could democratize AI agents the way Square simplified payments, making autonomous AI tools easier for businesses and individuals to adopt.

What This Means

The release of Goose highlights Block’s commitment to open-source AI development and expands the growing ecosystem of autonomous AI agents. While its first applications are engineering-focused, Goose’s extensibility suggests it could become a broader productivity tool for automation across industries. As more companies explore open-source AI frameworks, Goose could help bridge the gap between cutting-edge AI and real-world usability.

To use Goose, you can install it today here.

Editor’s Note: This article was created by Alicia Shapiro, CMO of AiNews.com, with writing, image, and idea-generation support from ChatGPT, an AI assistant. However, the final perspective and editorial choices are solely Alicia Shapiro’s. Special thanks to ChatGPT for assistance with research and editorial support in crafting this article.