Why Block handed Goose to the Linux Foundation

Key Concepts

• Goose: An open-source, model-agnostic AI coding agent originally developed by Block.
• Agentic AI Foundation (AIF): A newly formed foundation providing open governance for AI agent technologies.
• MCP (Model Context Protocol): A standard protocol for connecting AI models to data sources and tools.
• Open Governance: A framework that allows multiple companies to contribute to a project without concerns over single-entity trademark control.
• Mesh LLM: An experimental architecture for distributed, local compute sharing across multiple devices.
• Agent Client Protocol (ACP): A protocol enabling the decoupling of AI agent interfaces from their execution environments.

---

1. The Origin and Evolution of Goose

Goose began as an internal "skunk works" project at Block, designed to assist engineers with coding tasks. It gained rapid internal adoption, growing from a few users to thousands. Recognizing its potential, Block open-sourced the project in early 2025.

Despite its popularity, the project faced "headwinds" regarding growth because it lacked open governance. While permissively licensed (Apache), other companies were hesitant to contribute to a project where Block retained sole trademark control. This led to the creation of the Agentic AI Foundation, which combined Goose, MCP, and Agents MD into a single, neutrally governed ecosystem.

2. Technical Architecture and Capabilities

• Model Agnosticism: Unlike proprietary coding agents tied to specific models (e.g., Claude Code), Goose allows users to swap LLMs. It supports state-of-the-art models like Claude Opus 3.5, GPT-5.3, and GLM5.
• Task Fanning: Users can configure Goose to "fan out" a single task to multiple LLMs simultaneously to compare outputs and select the best result.
• Integration: Goose utilizes the Agent Client Protocol (ACP), allowing it to interface with IDEs like VS Code, IntelliJ, and Zed.
• Mesh LLM Experiment: A new, experimental feature that allows users to pool local compute resources across multiple laptops (similar to a decentralized "home-based" compute cluster) to handle tasks that exceed local capacity.

3. The Strategic Value of Open Source

Monik Certani argues that open-sourcing internal tools is not merely altruism but a sound business strategy:

• Maintenance Arbitrage: By open-sourcing core infrastructure (like the Android networking stack Block contributed in 2012), the company offloads long-term maintenance costs to the broader community.
• Talent Acquisition: A strong reputation in open source acts as a "flywheel," attracting top-tier engineering talent who prefer working in environments that contribute back to the ecosystem.
• Cultural DNA: Block’s history of building on open-source stacks (OS, databases, middleware) makes open-source contribution a natural extension of their development process.

4. Future Roadmap: Decoupling and Portability

The next 6–12 months for the AIF ecosystem will focus on:

• Deployment Flexibility: Moving away from laptop-only execution to a model where Goose can run in Docker containers, on remote servers, or in local meshes.
• Seamless Handoffs: Enabling users to start a task on a local machine, migrate the state to a server-side container, and resume the task on a mobile device without friction.
• Standardization: Certani positions Goose as the "Firefox of the agent world"—a reference implementation that keeps the industry honest regarding standards and interoperability.

5. Notable Quotes

• "The biggest headwind that we saw was lack of open governance... some companies were reticent to contribute back because we still own the trademarks." — Monik Certani on why the foundation was necessary.
• "I don't think it's altruism. I think it makes good business sense... guess how much we spend maintaining [our Android libraries] today? Absolutely zero. Everyone in the world is maintaining it for us." — Monik Certani on the economic benefits of open source.
• "I like to think of it as what we see in the browser space... I see [Goose] as the Firefox of the agent harness world where we get to be the reference implementation." — Monik Certani on the future of the agent ecosystem.

---

Synthesis

The transition of Goose from an internal Block tool to a cornerstone of the Agentic AI Foundation represents a shift toward standardized, model-agnostic AI development. By prioritizing open governance and decoupling the agent interface from the underlying LLM, the AIF aims to foster a pluralistic ecosystem rather than a "winner-takes-all" market. The core takeaway is that for AI agents to reach maturity, they must move beyond proprietary silos and adopt protocols (like MCP and ACP) that allow for portability, collaborative maintenance, and distributed compute.