Team Sync
Team Sync is the multi-agent control plane. Enroll company devices, share agent capabilities across teams, and roll them out across your fleet.
Team Sync is how the Company Brain extends from a single shared knowledge layer to a governed fleet of agents. It lets you enroll your company's machines as managed devices, package agent capabilities into reusable skills, share them across the company, and deploy them to your whole fleet at once.
Team Sync is the operational side of "one control plane for your agents." Connectors give every agent the same data; the Ontology Graph gives them the same understanding; Team Sync gives them the same capabilities, centrally managed.
The pieces#
The machines across your company where agents run. Each is enrolled once and then managed centrally. You can see its status and exactly which capabilities are installed on it.
The agents running on each device: Codex, Claude Code, Openclaw, Cursor, Hermes, or your own. Every agent connects to the same brain for knowledge and memory.
Reusable bundles that configure an agent's behavior. Build a skill once, share it to the company library, and deploy it everywhere it's needed.
Enrolling a device#
Enrolling a machine is a one-step action: generate an enrollment for it, run the provided command on the device, and it self-registers and stays connected to the brain. There's no manual configuration to maintain, and enrollments are single-use so a device can't be added by accident.
Sharing and deploying capabilities#
The lifecycle moves a capability from one person to the whole company:
Build (private) ──▶ Share to company library ──▶ Deploy to devices ──▶ Running on agents- Build a skill privately and test it.
- Share it to the company library so teammates can discover and reuse it. Re-sharing an updated version replaces the old one cleanly, with no duplicates.
- Deploy it to the devices and agents that need it, and watch the rollout complete. Each skill tracks where it's installed and flags drift if a device diverges from the deployed version.
Capabilities that improve from shared usage are built in, including agent self-learning and multi-agent learning across the fleet.
Why it matters for the enterprise#
- Consistency: every agent runs vetted, versioned capabilities instead of ad-hoc local configuration.
- Speed: push a new capability to the entire fleet at once.
- Visibility: a live inventory of which agents run which capabilities, with drift detection.
- Governance: capabilities are shared through a controlled company library, and every device and deployment action is auditable. See SSO & Audit.
How agents connect to and consume the brain.