AI Models & Platforms

SandboxAQ Makes Switch Free to Put Any AI Agent in Slack and Teams

mm
Add Unite.AI to your preferred sources on Google

SandboxAQ has released Switch, software that pulls AI agents built on rival frameworks into the same Slack, Microsoft Teams, or Discord channels where employees already work, and made the code publicly available at no cost. The company announced the launch on August 26, 2026, positioning Switch as the coordination layer between the agents developers are building and the collaboration tools enterprises have already paid for.

Switch turns a chat channel into what SandboxAQ calls a room: a shared workspace where people and agents operate with the same context, resources, and history rather than handing work between separate tools. Teams connect agents built with Claude Code, Google’s Agent Development Kit, LangChain, OpenAI’s tooling, and other frameworks, and the room keeps the context when participants join, leave, or hand off tasks. One agent can sit in multiple rooms at once, each carrying its own participants and rules. Connections run through the Model Context Protocol, APIs, and adapters, so the setup is vendor agnostic: models and agents can be swapped across providers without moving the workflow into a single AI ecosystem.

“Every team we talk to has capable agents trapped in separate tools, and a coordination tax eating the productivity those agents were supposed to deliver,” said Mohammed Aboul-Magd, General Manager at SandboxAQ, in the announcement. “If you can message a colleague, you can put an agent to work.”

Deployment runs through the Switch Console, a desktop application that sets the rest up, and the server component runs on infrastructure the customer controls, according to the Flint AI documentation. Two editions, Quick Start and Team, deploy in minutes, with an extensible architecture for enterprise deployment.

The Flint AI Product Line Behind Switch

Switch is the second product in Flint AI, SandboxAQ’s agent portfolio, and the release maps a deliberate funnel. The free Flint CLI, installed with a single pip command, scans and evaluates individual agents locally before they ship; the docs say nothing leaves the developer’s machine. Switch then takes agents that pass and puts them to work in shared rooms. A third layer, Flint Platform, is listed as coming soon on both the product page and the docs, promising discovery of agents in a company’s code, runtime session monitoring, and guardrail enforcement, delivered as a cloud service with GitHub integration and an SDK wrapper.

That sequencing tells buyers where the commercial product lives. The CLI and Switch are free and run on the customer’s own hardware; the platform layer, where governance and measurement sit, is the one SandboxAQ describes as cloud-based and unreleased. The Flint AI site also lists a weekly 20-minute live onboarding session for Switch run by its engineering team, starting August 26, 2026, alongside tutorials, a quickstart, and a Discord community.

SandboxAQ says it used Switch to build Switch, with its engineering, design, and marketing teams collaborating inside rooms throughout development. The company describes incident-response teams pulling specialized agents into a live channel that already holds the timeline, marketing rooms staffed with research, copy, and design agents, and legal teams adding drafting and review agents that carry a deal’s full context between rounds.

The License Free Isn’t Quite Open Source

SandboxAQ’s headline calls Switch open source, and the code is public. The actual license is Apache 2.0 with the Commons Clause attached, and the condition’s own authors are direct about what that means: “No,” it is not open source under the definition stewarded by the Open Source Initiative. The clause strips one right from Apache 2.0, the right to sell the software itself, defined as offering third parties, for a fee, a product or service whose value derives substantially from Switch’s functionality.

In practice the restriction lands on a narrow target. Enterprises can run Switch internally, modify it, redistribute it, and ship and charge for their own products built on top of it; SandboxAQ’s release says users can utilize the software within their enterprise, personally, and issue free products with it. What no one can do is take Switch, rebrand it, and sell it as a hosted service, the exact move a cloud provider or rival vendor would make. The pattern matches how infrastructure companies have defended commercial layers around free code for years, and it aligns with the Flint roadmap: the rooms are free, and the governance platform above them is the product.

For a company best known for Large Quantitative Models in drug discovery, catalysis, and navigation, and a recent run of AWS Marketplace and Claude integrations for its chemistry models, Switch is a different kind of bet: give away the coordination plumbing, get rooms running inside enterprises on any model, and let the paid platform arrive after the agents are already there. AWS made a similar infrastructure move this month, backing a federation layer for agent registries, as vendors race to own the connective tissue between agents rather than the agents themselves. The practical test starts now: the code is on GitHub, the onboarding sessions are weekly, and Flint Platform’s arrival will show what SandboxAQ charges for once the rooms fill up.

Aiden Cross is an AI-generated strategist at Unite.AI, covering AI product strategy, execution, and the practical challenges of turning experimental models into scalable, market-ready products. His work focuses on how startups and enterprise teams move from prototypes and demos to reliable systems used by real customers.

With a pragmatic and detail-oriented perspective, Aiden analyzes product roadmaps, go-to-market strategies, platform decisions, and organizational trade-offs that determine whether AI initiatives succeed or stall. He pays particular attention to deployment realities, user adoption, infrastructure constraints, and the alignment between technical capability and business value.

Articles authored by Aiden Cross are AI-generated and reviewed by Unite.AI’s editorial team to ensure clarity, accuracy, and responsible coverage of how AI products are built, shipped, and scaled in the real world.