You set the direction.
Your Robots get to work.

1. Meet Sam. Meet Sam. He writes books and promotes them with his marketing team. He also leads research into machine learning. Today, a chapter and a book launch need his attention. So do an LLM study and a European project budget. He keeps switching teams and searching for the right files. Even with AI chat, he has to explain the same context again. His own writing and research keep getting pushed aside.
2. WebChat and Copilot. Sam opens WebChat, the interface to the AssistOS Copilot. He describes what needs doing in each team's workspace. Books, Marketing, and Research each have a separate workspace. Each workspace has different Robots, roles, and access rules. Copilot assigns tasks to that workspace's RoboTeam and tracks their progress. Changing workspace brings Sam that team's Robots and colleagues.
3. Build the RoboTeam. Sam first sets up the Books RoboTeam. He chooses agents from the AssistOS catalog. These are software packages that run in Linux containers. His own IT team can also prepare custom agents. Each Robot is a configured instance of that software. Sam sets its subscriptions, preferences, permitted skills, and instructions. He also defines its role and workspace access. SCRIPTA drafts chapters, Proof reviews them, and Layout prepares the pages. Sam remains the author.
4. Background tasks. With his team ready, Sam asks Copilot to prepare a book preview. He also asks it to check the references. WebChat shows what each Robot is doing in the background. Robots can run CLI commands, coding agents, or custom agents. Robots can also use agents that check sources and results for trustworthy AI. Sam can redirect, pause, or stop a task from WebChat. He can then return to his work while the Robots continue.
5. Write and publish. The preview is ready, so Sam opens the book in Explorer. This is AssistOS as an IDE, an integrated development environment. Developers use one to keep their files, editor, and checks together. Sam does the same for his book. He compares SCRIPTA's edits with the original chapter. He reads Proof's notes before accepting the revision. When Sam approves the finished book, SCRIPTA publishes it on scriptahub.com.
6. Browser use. Sam switches to Marketing to help readers discover his book. Scout uses its browser to research relevant publications. Meanwhile, Studio prepares material for the book campaign. Sam can watch the browser, pause Scout, or take control himself. He reviews the outreach draft and decides whether to send it.
7. Computer use. Sam next opens the Research team's European project budget. The budget has its own AssistOS workspace. Hugo prepares staff, travel, and equipment costs in a desktop spreadsheet. This is computer use: a Robot working in an application Sam can watch. Sam can take control at any point. He reviews the figures with the project coordinator before submission.
8. Research. Sam then opens the workspace for the LLM study. Atlas compares baseline and smoothed LLM outputs. Vera evaluates the results. The team keeps its scripts, data, and notes together. Any workspace folder can stay local or use a GitHub repository. Here, the researchers use GitHub to review and share their changes.
9. WebMeet. Sam joins Lee and Robin in the study's WebMeet room. Atlas, Vera, and Hugo join as visible Robot participants. Lee asks Vera to add a held-out test to the shared plan. Everyone sees the change as they discuss it. The room uses the study's approved context. The meeting and its work stay with the right project.
10. Switch teams. After the meeting, Sam switches workspace and RoboTeam. Books has SCRIPTA, Proof, and Layout. Marketing has Scout, Studio, and Pulse. Research has Atlas, Vera, and Hugo. Each workspace keeps its own Robot settings, files, and meeting rooms. Colleagues with access can connect from a laptop or phone. They can manage Robots together and follow work across subprojects. Folders can stay local or connect to GitHub.
11. Why an OS? Linux containers help isolate agent code and protect the host machine. AssistOS can run on your computer or a cloud server shared by a team. Axiologic developed the open-source platform in the European Achilles project. The research project created an environment for controlling AI agents. Individual agents can have different licenses. This is the OS idea applied to virtual workers. AssistOS manages their homes, tasks, and permissions. Your project files and history remain when you change models or agents.
12. Results. Sam's day is just one example. Robots can help with many kinds of intellectual work. They can control browsers, use desktop apps, or run CLI tools. You give them those capabilities through their configured agents. The AssistOS team is happy to help you build your own integrations. AssistOS is open source and easy to install. Visit the Installation page to set up your first workspace. Open the Research menu for technical documentation. Thank you!

Ideas on the final board: review contracts, compare suppliers, check invoices, plan schedules, prepare courses, translate documents, clean datasets, analyse surveys, test software, run simulations, build dashboards, organise knowledge.

This film illustrates a future AssistOS workflow. The catalog, Robots and integrations shown have different levels of maturity. The charts and budget are illustrative.