Run your agents as one hive.

Work with agents across your computers, give them shared knowledge, track what they do, and keep approvals, apps, wallets, and setup in one place.

Find your path

Get oriented, connect your machines and runtimes, and learn the daily workflows for agents, memory, work, files, and integrations.

Follow the boundaries between the dashboard, local services, trusted machines, shared storage, and managed cloud features.

Turn requests into governed work, run autonomous companies, build and host apps, and verify outcomes before they count as complete.

Review the local-product and managed-service boundary, paid capabilities, public SDK, incentives, treasury design, and long-term product model.

See your trusted machines, choose the right runtime, talk to agents, and route work across the fleet.

Give every permitted agent shared memory, source knowledge, reusable skills, coordinated work, and setup-once credentials across trusted machines.

Set a company goal, assign a crew, define budgets and approval boundaries, and inspect evaluated outcomes and run history.

Build software with agents, preview it locally, publish it when ready, and send suitable workloads to available compute.

Understand agent wallets, governed spending, paid APIs, crypto and stock workflows, and the safeguards around money-moving actions.

Extend what agents can do with packaged skills, connected services, code proof, simulations, and runtime-specific guides.

01 Understand the product Start with the user guide and learn the core mental model before connecting more machines or runtimes.
02 Choose the workflow you need Jump to agents, work, memory, companies, apps, compute, wallets, trading, or integrations.
03 Keep the system healthy Use the troubleshooting cookbook when a machine, runtime, shared service, or integration does not behave as expected.