Recommended Free Tools
Yes, Block has AI agents—but not one single product called “Block’s AI agent.” The company’s strategy spans goose, a general-purpose agent that began as an internal tool; Square Managerbot, a proactive assistant for merchants; Moneybot, a Cash App financial assistant; and Cashbot, a Cash App customer-support agent.
The clearest customer-facing launch is Managerbot, which entered open beta for a broader group of Square sellers on April 28, 2026. Goose is the technical foundation of Block’s wider agent push, but it is not a new ChatGPT-like model: it is an open-source software layer that connects foundation models to tools, data and workflows.
Block’s AI products are aimed at different users
| Product | Primary audience | What is established | Availability |
|---|---|---|---|
| goose | Developers, technical teams, researchers and advanced users | General-purpose, tool-using agent for research, writing, coding, automation, analysis and repeatable workflows | Public open-source desktop app, CLI and API |
| Managerbot | Square sellers | Monitors operations, surfaces insights, automates routine tasks and can perform specified operational actions | Open beta for a broader group of sellers announced April 28, 2026; eligibility can vary |
| Moneybot | Cash App customers | Proactive financial assistant and an early example of Block’s “proactive intelligence” strategy | Block described it as moving toward broader availability; a complete public feature list is not established here |
| Cashbot | Cash App customers seeking support | AI-powered customer-support agent | Rolled out in the second quarter of 2025 |
These are separate experiences. Installing goose does not grant access to Square seller records or Cash App data, and Managerbot is not a general-purpose coding or research agent.
Managerbot is Block’s most visible business agent
Managerbot lives inside Square Dashboard. Square says it can watch operational information, identify useful insights, automate routine work and help protect a seller’s business. Its proactive design matters: it is intended to surface issues and next steps rather than wait for a perfectly phrased question.
Outdated Drivers Are Slowing You Down
One free scan finds every outdated or missing driver and matches the right update for your exact hardware.Free scan · exact hardware matchPC Slower Than It Used to Be?
A free scan shows the junk files, broken settings and background clutter dragging Windows down - then fixes them in one click.Free scan · Windows 10 & 11#1 Best Overall
Square’s launch announcement gives a concrete example: a seller can use a voice or text command to update item availability across multiple locations. That is materially different from a chatbot that only explains how to change inventory settings. The public announcement establishes bounded task execution, not unrestricted autonomy or error-free decisions about staffing, purchasing, finances or compliance. Square’s open-beta announcement also means access may depend on geography, account configuration and the Square products a seller uses.
Goose is the general-purpose agent behind the strategy
Block developed goose in early 2024 so foundation models could execute work across the enterprise instead of merely answering questions. The project is now public and has moved from Block’s GitHub organization to the Agentic AI Foundation at the Linux Foundation. It is accurate to call goose Block-created or originally developed by Block, but not to describe it as a Block-only project today.
Goose is an agentic software layer, or harness. It can use models from multiple providers and connect them to tools through the Model Context Protocol (MCP). The public project documents desktop, command-line and API forms, extensions for services such as databases, browsers, GitHub and Google Drive, reusable recipes, subagents and permission controls. Supported providers include Anthropic, OpenAI, Google, Ollama, OpenRouter, Azure OpenAI and Amazon Bedrock. Block has not announced a standalone frontier model that replaces those providers.
Installing and configuring goose
The project publishes this CLI installation command:
What’s actually slowing this PC down?
Pick the symptom - the matching free tool is one click away.
Rank #2
curl -fsSL https://github.com/aaif-goose/goose/releases/download/stable/download_cli.sh | bashDesktop builds are listed for macOS, Linux and Windows. After installation, configure a model provider in the desktop app by following:
- Open the sidebar and select Settings.
- Open the Models tab.
- Select Configure providers.
- Choose a provider and enter its credentials.
The documented CLI alternative is:
goose configureGoose is released under the Apache 2.0 license. The software itself is open source, but model inference, hosted services and connected tools can carry separate usage costs. Provider details are documented in the goose provider guide.
Moneybot and Cashbot serve different Cash App jobs
Moneybot
Moneybot is Cash App’s proactive financial assistant. Block’s first-quarter 2026 shareholder materials present it as an example of “proactive intelligence”: the system can bring relevant information or possible actions to a customer instead of waiting for an open-ended request. The available evidence does not establish a universal feature list, universal access or that Moneybot is a regulated financial adviser.
Cashbot
Cashbot is support-focused. Block’s prepared remarks identify it as an AI-powered customer-support agent rolled out in the second quarter of 2025. It should not be conflated with Moneybot’s financial-assistance role or with Managerbot’s merchant operations.
Do these 3 things before closing this tab:
1Repair Windows errors before they cause bigger problems2Scan for outdated or missing drivers - takes under a minute3Clear out junk files and repair common Windows errorsWhy Block wants agents inside its products
Block controls valuable operational contexts: payments, commerce, seller activity and consumer financial accounts. An agent embedded in those systems can potentially turn data into an action—such as correcting an item’s availability—without making the user move information between several apps.
Block also frames agents as an internal productivity strategy. At its 2025 Investor Day, the company said more than 6,500 of its 10,000-plus employees used goose weekly. Block reported that engineers saved eight to 10 hours per week and reduced manual work by approximately 25%; those are company-reported figures, not independent measurements. In first-quarter 2026 materials, Block said production code changes per engineer were more than 2.5 times the January level by mid-April and that incident rates after production code changes fell substantially year over year. Those figures do not prove that agents caused all of the improvement.
Block is also extending Square’s reach into other AI interfaces. On July 1, 2026, Square announced integrations with ChatGPT and Claude intended to help sellers be discovered and, in supported experiences, transact through AI-powered conversations. The strategic aim is less about owning the underlying language model than about owning the payments, commerce and action layer behind an AI-assisted transaction.
How the AI strategy relates to Block’s layoffs
The Associated Press reported on February 27, 2026, that Block cut approximately 4,000 jobs, or about 40% of its workforce. Jack Dorsey linked the reduction to changes in how the company could operate with AI-enabled productivity.
That is important context, but it is not proof that AI directly replaced 4,000 individual workers. The reported workforce reduction, Block’s explanation and the precise number of jobs attributable to AI are different claims. A responsible reading is that Block presented AI-driven operating leverage as part of the rationale for a major restructuring; the independent causal breakdown remains unestablished.
Independent reader supportYour contribution helps us test, update, and keep practical guides available for everyone.What users can and cannot do today
Goose is best suited to technical, configurable work
- Good fit: local or self-managed agents, MCP-connected tools, multi-provider model choice and workflows beyond code completion.
- Poor fit: a turnkey consumer chatbot, a no-configuration experience, or an organization unwilling to manage credentials and tool permissions.
Goose documentation describes prompt-injection detection, tool permissions, sandbox mode and an adversary reviewer. These are safeguards, not guarantees. Giving an agent access to files, repositories, databases or external services creates risks including prompt injection, data leakage, destructive commands and provider-retention concerns.
Managerbot is best suited to Square-based operations
- Good fit: an existing Square seller that wants insights and routine actions based on data already in Square.
- Poor fit: a business outside the Square ecosystem, a company needing deep platform-neutral customization, or a mission-critical process that requires guaranteed human judgment.
Because Managerbot was in open beta, sellers should treat its availability and behavior as account- and market-dependent rather than as a universal production guarantee.
Cash App assistants require extra caution
Moneybot and Cashbot operate in a financial-services context. Eligibility, jurisdiction, account status and regulatory restrictions may affect access. Do not assume that Moneybot provides personalized regulated advice or that either Cash App assistant is available to every customer.
Best Value
How Block’s agents compare with other AI agents
| Category | Examples | Defining characteristic |
|---|---|---|
| General assistants | ChatGPT, Claude, Gemini | Broad conversation and tool use across many tasks |
| Coding agents | Claude Code, Codex, Cursor and similar tools | Developer-focused repository, testing and implementation workflows |
| Open-source or local agents | Goose and other MCP-compatible projects | More control over models, deployment and extensions, with greater setup responsibility |
| Embedded business agents | Square Managerbot and assistants in commerce, CRM or accounting platforms | Access to proprietary operational data and the ability to perform platform actions |
| Customer-support agents | Cashbot and comparable support systems | Resolve or route customer issues within a controlled service environment |
Goose therefore competes mainly at the agent-framework and workflow layer, not as a single model against every model vendor. Managerbot competes more directly with business-platform assistants because its advantage is Square’s operational context.
Bottom line: Block is building an agentic operating layer
Block clearly has AI agents, but the important story is the portfolio. Goose supplies a flexible, tool-using foundation for internal and open-source work; Managerbot puts proactive automation in front of Square sellers; Moneybot and Cashbot address consumer finance and support.
Block’s distinctive bet is to place agents where money and commerce already move. Whether that becomes durable differentiation will depend on reliability, permissioning and user trust—especially when an agent can act on financial or business data—not simply on how fluent its underlying model sounds.
Quick Recap
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
Quick wins for a faster PC:
Scan for outdated or missing drivers - takes under a minuteDriver Scan →Clear out junk files and repair common Windows errorsFree Scan →
