Goose Tutorial: Extensible Open-Source AI Agent for Real Engineering Work

Learn how to use block/goose to automate coding workflows with controlled tool execution, strong provider flexibility, and production-ready operations.

Why This Track Matters

Goose is one of the highest-velocity open-source coding agents and combines desktop + CLI interfaces with deep MCP extension support.

This track focuses on:

• setting up Goose quickly across desktop and terminal workflows
• understanding the agent loop, tool execution, and context management model
• controlling permissions, risk boundaries, and extension behavior
• operating Goose safely for teams and production usage

Current Snapshot (auto-updated)

• repository: block/goose
• stars: about 49.4k
• GitHub release reference: v1.37.0 (checked 2026-06-15; release metadata on GitHub)

Mental Model

flowchart LR
    A[Developer Request] --> B[Goose Interface UI or CLI]
    B --> C[Agent Loop]
    C --> D[Tool or Extension Calls]
    D --> E[Context Revision and Summarization]
    E --> F[Actionable Output or Code Changes]

Chapter Guide

Chapter	Key Question	Outcome
01 - Getting Started	How do I install and launch Goose quickly?	Working baseline setup
02 - Architecture and Agent Loop	How does Goose process requests and tool calls?	Strong runtime mental model
03 - Providers and Model Routing	How do I choose and configure providers safely?	Reliable model configuration strategy
04 - Permissions and Tool Governance	How do I control automation risk in daily usage?	Safer execution boundaries
05 - Sessions and Context Management	How does Goose sustain long-running sessions?	Durable conversation and token management
06 - Extensions and MCP Integration	How do I add capabilities with built-in and custom MCP servers?	Extensible workflow design
07 - CLI Workflows and Automation	How do I automate Goose in scripts and CI-like flows?	Repeatable command-driven workflows
08 - Production Operations and Security	How do teams run Goose in production responsibly?	Governance and operations runbook

What You Will Learn

• how to use Goose across desktop and CLI surfaces
• how the Goose agent loop uses tools, extensions, and context revision
• how to balance autonomy and safety with permission modes and tool controls
• how to operationalize Goose for reproducible, team-scale engineering workflows

Source References

• Goose Repository
• Goose Quickstart
• Install goose
• Supported providers
• Using extensions
• CLI commands
• Permissions guide
• Architecture guide

Related Tutorials

• OpenCode Tutorial
• Cline Tutorial
• Roo Code Tutorial
• Crush Tutorial

---

Start with Chapter 1: Getting Started.

Navigation & Backlinks

• Start Here: Chapter 1: Getting Started
• Back to Main Catalog
• Browse A-Z Tutorial Directory
• Search by Intent
• Explore Category Hubs

Full Chapter Map

1. Chapter 1: Getting Started
2. Chapter 2: Architecture and Agent Loop
3. Chapter 3: Providers and Model Routing
4. Chapter 4: Permissions and Tool Governance
5. Chapter 5: Sessions and Context Management
6. Chapter 6: Extensions and MCP Integration
7. Chapter 7: CLI Workflows and Automation
8. Chapter 8: Production Operations and Security

Generated by AI Codebase Knowledge Builder