@Fluyeporlaweb: Jack Dorsey built the most COMPLETE local AI agent that exists. Then he donated it to the Linux Foundation so that no o…

X AI KOLs Timeline Tools
local-ai-agent open-source jack-dorsey linux-foundation goose developer-tools ai-agent

Summary

Jack Dorsey built and open-sourced Goose, a comprehensive local AI agent that autonomously performs development tasks like installing dependencies, running code, testing, and deploying. Donated to the Linux Foundation, it supports multiple LLMs and includes features like parallel subagents, MCP extensions, and adversary mode.

Jack Dorsey built the most COMPLETE local AI agent that exists. Then he donated it to the Linux Foundation so that no one controls it. 46.4k stars. 518 contributors. 137 releases. Updated a few hours ago. It's called Goose and it's the closest thing I've seen to a real AI agent running on your machine. It doesn't just suggest code. It installs dependencies, runs, edits, tests, debugs, and deploys. On its own. Without you watching over it. What sets it apart from Claude Code or Codex: Native desktop app + CLI + API - choose how to use it Works with any LLM: Claude, GPT, Gemini, DeepSeek, Ollama and 15 more Uses your existing subscriptions - no paying for more APIs 70+ MCP extensions: GitHub, Google Drive, databases, browser and more Parallel subagents - breaks down complex tasks and runs them in parallel Recipes: save workflows as YAML and share them with your team Built-in adversary mode - a reviewer that detects prompt injections and unsafe actions Compatible with Claude Code and Codex as providers via ACP Native Rust. Mac, Linux and Windows. Apache 2.0. The part that blew my mind the most: You can connect it to Claude Code or Codex as a subagent. Goose orchestrates. They execute. The agent that coordinates other agents. the link
Original Article
View Cached Full Text

Cached at: 06/08/26, 07:21 AM

Jack Dorsey built the most COMPLETE local AI agent that exists.

Then he donated it to the Linux Foundation so that no one controls it.

46.4k stars. 518 contributors. 137 releases. Updated a few hours ago.

It’s called Goose and it’s the closest thing I’ve seen to a real AI agent running on your machine.

It doesn’t just suggest code.

It installs dependencies, runs, edits, tests, debugs, and deploys.

On its own. Without you watching over it.

What sets it apart from Claude Code or Codex:

Native desktop app + CLI + API - choose how to use it Works with any LLM: Claude, GPT, Gemini, DeepSeek, Ollama and 15 more Uses your existing subscriptions - no paying for more APIs 70+ MCP extensions: GitHub, Google Drive, databases, browser and more Parallel subagents - breaks down complex tasks and runs them in parallel Recipes: save workflows as YAML and share them with your team Built-in adversary mode - a reviewer that detects prompt injections and unsafe actions Compatible with Claude Code and Codex as providers via ACP Native Rust. Mac, Linux and Windows. Apache 2.0.

The part that blew my mind the most:

You can connect it to Claude Code or Codex as a subagent.

Goose orchestrates. They execute.

The agent that coordinates other agents.

the link

Similar Articles

aaif-goose/goose

GitHub Trending (daily)

Goose is an open-source, extensible AI agent capable of installing, executing, editing, and testing code using any LLM.