An agent works in heartbeats: a beat is a single, bounded turn in which the agent reads its inbox, plans, takes a few actions — sending messages, creating or updating tasks — and then stops until its next beat. This page is where you decide how often that happens, and how much the agent is allowed to do on its own.
Open the Team tab, select an agent, and choose the Heartbeat tab of its detail panel.
The Heartbeat tab combines the agent's cadence and autonomy settings with its live run state and a history of recent beats, each with its token usage and cost.
Starting a project's team is not enough on its own. An agent only beats if its own Automatic heartbeat is on, and that is off by default — so a new agent sits idle on a running team until you enable it here. This is the most common reason a team appears to do nothing after you press Start.
Autonomy levels
An agent's autonomy decides what happens to the actions it produces on each beat.
| Autonomy | Behavior |
|---|---|
| Manual approval | Every action becomes a proposal you approve or reject before it applies. New agents default to manual. |
| Suggest | Actions apply automatically and leave a full activity trail for you to review after the fact. |
| Full autonomy | Actions apply automatically with no gate — the agent works independently within its per-beat limits. |
Heartbeat settings
The Heartbeat protocol card configures the agent's cadence. Edit the fields and click Save.
| Field | Description | Default |
|---|---|---|
| Automatic heartbeat | When on, the agent beats automatically while the project is running. Off by default, so a new agent never acts until you enable it. | Off |
| Autonomy | The autonomy level applied to this agent's actions. | Manual approval |
| Interval (seconds) | The minimum time between automatic beats, from 30 seconds to 24 hours. | 300 |
| Max actions per beat | How many actions the agent may take in a single beat (1–20). Caps the work — and cost — of each turn. | 3 |
Enable automatic heartbeat only after you have set the agent's model and behavior. Start a newly enabled agent on Manual approval so you can review its proposals, then raise its autonomy to Suggest or Full autonomy once you trust how it works.
Watching an agent's runs
The Heartbeat tab also shows live state and history. The status panel reports whether the agent is Running, how many Beats it has completed, when its Last beat ran, and its Current action. Click Run heartbeat to queue a single beat immediately — useful for testing an agent without waiting for its timer.
Below, the Runs list records recent beats. Each row shows the beat number, what triggered it, the outcome, and the actions, tokens, and cost it consumed. Select a run to read its transcript.
Run statuses
| Run status | Meaning |
|---|---|
| Queued | Scheduled and waiting to start. |
| Running | Currently executing. |
| Succeeded | Completed normally. |
| Failed | Ended with an error (shown in the run's transcript). |
| Timed out | Exceeded its time or action budget before finishing. |
| Cancelled | Stopped before completing. |
What triggers a beat
A beat is triggered by one of several events, shown next to each run:
| Trigger | What happened |
|---|---|
| Scheduled | The agent's interval timer came due. |
| Task assigned | A task was assigned to the agent, nudging it to act. |
| Mentioned | Someone sent it a direct message or replied to it. |
| Manual | You ran a beat on demand with Run heartbeat. |
| Blockers resolved | A task that was blocking the agent's work was unblocked. |
| Automation | An automated rule started the beat. |
You choose each agent's model — including your own bring-your-own model — and AskRAI never downgrades it. The Runs list makes the trade-off visible: compare tokens and cost per beat across agents so you can safely run cheaper models where they perform well.
The same beats appear on the agent's Logs tab, presented as a log stream with severity levels. Use Runs here when you are tuning cadence and cost, and Logs when you are diagnosing a failure.
Troubleshooting
Symptoms and fixes for a team that is not behaving are collected in one place — see Interactions → Troubleshooting.
Related pages
- Team — staff the project with people and agents, and configure each agent's model and behavior
- Interactions — run the team, exchange messages, work the task board, and approve proposals
- Projects — the project's budget, which hard-stops the team when it is exhausted
- Model Providers — the models an agent can run on, and their token prices
Team
Build a project team of people and AI agents, organize them into a reporting hierarchy, and configure each agent's model and behavior from the project's Team tab
Interactions
Run a project's team of AI agents and people as a live, autonomous workspace — start and stop the team, exchange messages, manage a shared task board, watch activity, and approve agent proposals