Set up a role
Choose the agent’s task, capabilities, and limits.
Loading...
Platform
Organize agents, assign work, and review the actions and results they produce.
Identity & AI agents
Agent workforce management brings identity, tasks, tool access, and recorded execution into the same operational picture. Your team stays responsible for the decisions that matter.
External integrations require their own configuration and credentials. Autonomous posting or outreach should be enabled deliberately.
Platform
Choose the agent’s task, capabilities, and limits.
Run tasks through configured tools and workflows.
Use execution records and receipts to inspect what happened.
Details and limitations
Describe the agents your team intends to operate, their ownership and the pool or resource grouping they belong to. The Console provides a place to inspect these records and their lifecycle. Give each entry a clear operational purpose so it can be distinguished from a test fixture.
Workspace records support oversight of configuration, status and administrative history. They help a team see which agent was registered, paused or terminated and who was responsible for the change. Those records should be compared with telemetry from the process that actually executes work.
Registering or resuming an agent record does not currently provision or start a native worker process. Native process provisioning and cancellation remain unfinished. Capacity records and administrative controls should not be presented as proof that a worker is running or that a monetary budget has been enforced.
Test record creation and lifecycle permissions separately from runtime execution. Before relying on automated operation, demonstrate where a worker starts, how it is stopped and what happens to in-flight work. Use a documented external runtime where required and retain its own operational evidence.