# Goose

Open-source desktop, CLI, and API agent from Block for coding and everyday workflows.

## Summary
Goose is Block's open-source AI agent for working across code, local workflows, tools, and extensions through desktop, CLI, and API surfaces.


## Guide
Goose is Block's open-source AI agent for working across code, local workflows, tools, and extensions through desktop, CLI, and API surfaces.

### What it is
Goose is an open AI agents resource tracked by OpenAgent.bot because it gives builders a concrete implementation path rather than just a product claim.

### Why it matters
Goose matters because it treats agent work as something that belongs on the user's machine, connected to local files, development tools, and extensions rather than trapped inside a chat tab.

### How it works
Start from the official repository or documentation, verify the license and runtime requirements, then test it on a narrow workflow before expanding it into production use.


## Use Cases
- Developer task execution: Use Goose to inspect repos, run tools, and work through coding tasks locally.
- MCP-backed workflows: Connect Goose to extensions or servers that expose tools to the agent.
- Desktop agent evaluation: Compare what a desktop-first agent can do compared with a browser-only chat UI.

## Alternatives
- Goose fits local developer workflows vs hosted agent products: Hosted agents can be convenient, but Goose is more interesting when local files, tools, and extension control matter.

### Getting Started
- Review the GitHub repository: https://github.com/block/goose
- Official source: https://block.github.io/goose/docs/
- Official source: https://block.github.io/goose/

### FAQ
- Is Goose open source?
  - Goose is listed with Apache-2.0 based on its official source links. Always re-check the repository or model card before production use.
- Who should evaluate Goose?
  - Developers who want a local agent for coding and workflow tasks
## Why It Matters
Goose matters because it treats agent work as something that belongs on the user's machine, connected to local files, development tools, and extensions rather than trapped inside a chat tab.


## Best For
- Developers who want a local agent for coding and workflow tasks
- Teams evaluating desktop and CLI agent UX
- Builders interested in MCP-style extension ecosystems

## Not For
- Users who want a fully managed consumer product with no setup work
- Teams that cannot review the linked source, license, and operational requirements before adoption

## What It Actually Does
- Local agent surface: Goose provides desktop and CLI paths for agent work.
  - Why it matters: Local surfaces are often more natural for coding, files, and repeatable workflows.
- Extension-oriented design: Goose connects agents to tools through extensions and provider configuration.
  - Why it matters: Agents become more useful when they can safely reach the tools a user already relies on.
- Permissive open-source release: Block has described Goose as Apache-2.0 open source.
  - Why it matters: A permissive license makes it easier to study and adapt the agent runtime.

## Typical Use Cases
- Developer task execution: Use Goose to inspect repos, run tools, and work through coding tasks locally.
- MCP-backed workflows: Connect Goose to extensions or servers that expose tools to the agent.
- Desktop agent evaluation: Compare what a desktop-first agent can do compared with a browser-only chat UI.

## How It Compares
- Goose fits local developer workflows vs hosted agent products: Hosted agents can be convenient, but Goose is more interesting when local files, tools, and extension control matter.

## Command Line
### Clone the Goose repository
Start from the official Block repository and follow the current install instructions in the docs.

```bash
git clone https://github.com/block/goose.git
```

## Facts
- Category: agents
- Resource type: agent
- Open source: yes
- License: Apache-2.0
- Last verified: 2026-04-19
- GitHub repo: block/goose

## Capabilities
- workflow-orchestration
- tool-calling
- mcp

## Structured Use Case Tags
- self-hosted-ai
- developer-workflow

## Getting Started
- Review the GitHub repository: https://github.com/block/goose
- Official source: https://block.github.io/goose/docs/
- Official source: https://block.github.io/goose/

## Links
- GitHub: https://github.com/block/goose
- Homepage: https://block.github.io/goose/
- Docs: https://block.github.io/goose/docs/
- Source: https://block.xyz/inside/block-open-source-introduces-codename-goose

## Structured Outputs
- JSON: https://www.openagent.bot/agents/goose.json
- Markdown: https://www.openagent.bot/agents/goose.md
- Canonical: https://www.openagent.bot/agents/goose