goose - AI Code Assistants Tool

Overview

Goose is an open-source, on‑machine AI agent for developers that automates end‑to‑end engineering tasks: it can scaffold projects, write and execute code, run tests, debug failures, and orchestrate multi‑step workflows. The project is designed to be provider‑agnostic (works with OpenAI, Anthropic, Gemini, local models, and many others), supports multi‑model configurations to balance quality and cost, and exposes both a CLI and a Desktop app so it can be embedded into personal developer workflows or CI pipelines. ([github.com](https://github.com/block/goose)) Goose emphasizes extensibility and integrations: it supports extensions (e.g., a Computer Controller for browser/file automation), integrates with external services via MCP (Model Context Protocol) servers, and can be extended with custom providers and plugins to read repos, Jira, Google Drive, Slack, and other tooling. The project is actively developed and maintained with frequent releases and community contributions — the GitHub repository lists an active release cadence (v1.26.1, Feb 27, 2026) and extensive docs/quickstart guides for getting started. Goose has been covered in technology press as part of Block’s open‑source agent strategy and as a participant in emerging agent interoperability efforts. ([block.github.io](https://block.github.io/goose/docs/getting-started/installation))

GitHub Statistics

• Stars: 32,190
• Forks: 2,928
• Contributors: 378
• License: Apache-2.0
• Primary Language: Rust
• Last Updated: 2026-03-02T20:29:05Z
• Latest Release: v1.26.1

Goose is highly active and widely adopted on GitHub: the repository shows 32.2k stars and ~2.9k forks, is licensed under Apache‑2.0, and has a multi‑release history (v1.26.1 listed Feb 27, 2026). The codebase is Rust‑based (Cargo.toml, rust toolchain files present) and contains a large number of commits and releases, indicating ongoing feature work and maintenance. Contribution and issue activity is substantial (open issues, frequent PRs, and an active community/Discord), which suggests good community health but also ongoing work on stability and UX. For docs and pragmatic setup instructions (desktop + CLI + Homebrew/download script), see the official docs. ([github.com](https://github.com/block/goose))

Installation

Install via brew:

curl -fsSL https://github.com/block/goose/releases/download/stable/download_cli.sh | bash

brew install --cask block-goose

brew install block-goose-cli

Invoke-WebRequest -Uri "https://raw.githubusercontent.com/block/goose/main/download_cli.ps1" -OutFile "download_cli.ps1"; .\download_cli.ps1  (Windows PowerShell)

Key Features

• Autonomously scaffold, implement, and run full projects from a single prompt.
• Write and execute code locally, then run tests and iterate on failures.
• Pluggable provider support — configure OpenAI, Anthropic, Gemini, local models, and routers.
• Extensions system (e.g., Computer Controller) for browser, file, and automation tasks.
• MCP integrations to connect GitHub, Google Drive, Jira, Slack and other context sources.
• Available as Desktop app and CLI; Homebrew and scripted installers for macOS/Linux/Windows.

Community

The Goose project has a large, active community and governance around responsible agent use. GitHub shows ~32.2k stars and ~2.9k forks with frequent releases (v1.26.1 on Feb 27, 2026) and active issue/PR traffic; the project maintains documentation, a Discord, and monthly contributor spotlights. Press coverage and third‑party reviews highlight both strong capabilities and caveats (model quality varies with provider; setup requires developer familiarity). Overall community signals point to rapid adoption and healthy contributor engagement, with ongoing work on stability, integrations, and standardization (Model Context Protocol / Agentic AI efforts). ([github.com](https://github.com/block/goose))

Key Information

• Category: Code Assistants
• Type: AI Code Assistants Tool