AI Content Agents do not have to wait for you to press run. You can set an agent to work on a recurring schedule, react to a change in a project, or hand off to another agent. This frees you from routine dispatch and keeps work moving when you are away from the app.

You control when an agent runs from its Triggers settings. There are three kinds of trigger: schedule, event, and manual. You can combine them on a single agent.

Use a schedule when you want the agent to run on a recurring cadence, for example every weekday morning or once a week.

Open the agent, go to its Triggers settings, and add a schedule. Set the cadence you want the run to repeat on. The agent then runs on its own at each interval, with no action from you.

Use an event trigger when you want the agent to react to work happening in a project. The available event is Project stage changed, which fires on every stage transition in the workspace the agent is scoped to.

Add the event trigger from the agent's Triggers settings. Because the trigger fires on all stage transitions, put the check for the specific stage inside the agent instructions. For example, tell the agent to act only when a project has reached the stage you care about, and to stop otherwise. The trigger starts the run, and the instructions decide whether there is work to do.

Use a manual trigger when you want to stay in control and run the agent yourself. Open the agent thread and start a run on demand. This suits one-off work and testing a new agent before you put it on a schedule.

An agent can also be triggered by another agent. Use this to chain work, so one agent finishes a step and hands off to a second agent for the next step. Set the triggering relationship from the agent's Triggers settings.

Set the agent's Scope to a workspace so it only fires for that workspace's projects. This matters most for the stage-change event, which would otherwise react to transitions you did not intend. Scoping keeps a scheduled or event-driven agent focused on the projects you want it to touch.

The agent thread shows each run and the tool calls it made. After a scheduled or triggered run, open the thread to confirm the agent fired when you expected and did what you wanted. If a run did nothing, the thread tells you whether the trigger fired and the instructions chose to stop, or the trigger did not fire at all.

Does the stage-changed trigger fire only for one stage?

No. It fires on every stage transition in the scoped workspace. To act on one stage only, add that check to the agent instructions.

Can two agents chain together?

Yes. One agent can trigger another, so you can pass work from one agent to the next.

If you need help, message us in the Relato Slack community or email [email protected].