BREAKING: Block (Jack Dorsey's company) just open-sourced a local AI agent

Sukh Sroay · tweet · 2026-04-10 source ↗ Reader 🖥️

Agentic AI Coding Agents Local AI MCP (Model Context Protocol)

Block has open-sourced Goose, a fully local, LLM-agnostic AI agent that autonomously executes entire engineering workflows rather than merely suggesting code, and this represents what 'agentic AI' actually looks like.

practical · promotional Twitter thread — A tech-hype announcement thread presenting Block's open-source AI agent Goose. The form is characteristic of tech influencer marketing: breaking-news framing, feature bullet lists, social proof via star count, and a call-to-action to follow/retweet.

Structure

Part	Role
Breaking announcement	Hook: establishes newsworthiness and names the actor (Block / Jack Dorsey)
Differentiation claim	Positions Goose as qualitatively different from coding assistants by emphasizing autonomy
Feature list	Provides evidence for the differentiation claim via concrete capabilities
Execution vs. suggestion contrast	Rhetorical climax: redefines 'agentic AI' as execution, not suggestion
Social proof and CTA	Legitimizes via GitHub stars, Apache 2.0 license, and Block's brand; then pivots to self-promotion

Key terms

local AI agent Coding Agents
An AI system that runs entirely on the user's machine, never sending code to external servers
agentic AI Agentic AI
AI that autonomously plans, executes, debugs, and delivers complete workflows without human intervention at each step
coding assistant Coding Agents
A tool limited to suggesting code snippets that a human must review, copy, and integrate manually
autonomous engineering agent Coding Agents
A system that independently builds projects, writes/runs code, debugs failures, and orchestrates workflows end-to-end
MCP server integration MCP (Model Context Protocol)
Native support for the Model Context Protocol, enabling extensibility and tool use

Propositions

explicitBlock has open-sourced a local AI agent called Gooseconf: high
explicitGoose is fundamentally different from coding assistants because it executes autonomously rather than suggestingconf: high
explicit⚡ disputed & corroboratedGoose works with any LLM and runs fully locally so code never leaves the user's machineconf: high
supports ←Google's AlphaEvolve Reaches General Availability with Evolutionary Code Optimization as a Service · p3 — Different architecture (cloud API generating candidates vs. fully local execution) but the same underlying principle: keep the customer's code/environment off the vendor's infrastructure as the enabling condition for enterprise trust.
refines ←Google Brings MCP Support to Colab, Enabling Cloud Execution for AI Agents · p1 — Qualifies the fully-local-agent virtue Goose is sold on by naming its costs: limited compute and the security risk of executing untrusted code locally.
contradicts ←Google Brings MCP Support to Colab, Enabling Cloud Execution for AI Agents · p2 — Goose's value is fully-local execution so code never leaves the machine; Colab argues local execution's compute and security limits are best solved by delegating to the cloud.
explicitAutonomous execution without human intervention is the defining characteristic of real agentic AIconf: medium
implicitGoose's popularity (33,500+ GitHub stars) validates its valueconf: medium
implicitBeing backed by Block (a major public company) makes Goose more credible and likely to be maintainedconf: medium
implicitcontestedExisting AI coding tools are inadequate because they require too much human involvementconf: high
contradicts ←Breaking: Autonomous Agents are a Shitshow · p6 — The Goose pitch treats human involvement as the flaw to eliminate; Marcus's vulnerability data argues unsupervised autonomy is exactly what is reckless.

Arguments

Developers should adopt Goose as their AI development tool strength: medium

Explicit premises:

Goose is 100% open source under Apache 2.0
It runs fully locally so code never leaves your machine
It works with any LLM
It autonomously builds, debugs, and ships code
It has 33,500+ GitHub stars and is built by Block

Implicit premises:

Open-source tools are preferable to proprietary ones
Data privacy (local execution) is important to developers
LLM-agnosticism is desirable because it avoids vendor lock-in
GitHub star count is a reliable proxy for tool quality
A tool backed by a large company is more trustworthy and sustainable

Inferences:

Because Goose executes rather than suggests, it saves developer time and effort
Because it is open source and local, it addresses the two biggest concerns developers have with AI tools (privacy and lock-in)

Goose represents what 'agentic AI' actually looks like strength: low