Hand a step to an AI Agent that decides what to do within the approvals and limits you set: allowed actions, a confidence threshold, and automatic spending limits.
A reasoning step inside an otherwise deterministic flow.
Most steps do exactly one fixed thing. An AI Agent step is different: it hands the record to a model that looks at the context and decides which of its allowed actions to take. You do not script the decision; you set the boundaries, and the agent reasons within them.
Use an agent step when the right move depends on judgement a fixed rule cannot capture, like drafting a reply that fits the thread or deciding which of several follow-ups suits a given deal.
Every agent step is bounded by the same set of controls.
| Field | Description |
|---|---|
| Name* | A label for the agent step so its purpose is clear in the flow and the run history. |
| Model | How hard the agent thinks: Light Tasks (the default), Standard, Advanced, or Advanced reasoning. Higher tiers cost more per run. See the Model section below. |
| Allowed actions | The exact set of CRM actions this agent may take. The agent can never act outside this list. |
| Autonomy | Whether the agent’s proposed actions wait for a human. See the autonomy dial below. |
| Confidence threshold | How sure the agent must be before an action is eligible to run on its own. Anything below the bar is sent for approval. |
| Spending limits | Automatic. Agent usage counts against your plan’s AI allowance, and each agent is limited to a share of it (25% per month, 5% per day) so a runaway agent is caught quickly. Nothing to configure. |
Choose how hard the agent thinks.
Each agent has a Model setting with four tiers:
The setting applies on every way the agent runs — scheduled, triggered, or run manually — so what you pick is what the agent uses everywhere.
How much the agent may do unattended.
The confidence threshold works alongside the mode: even an unattended action only runs without review when the agent’s confidence clears the bar you set. Below it, the action is routed to the queue.
A preset agent that drafts two reply options to your inbound email — you pick one and send.
The Inbox Reviewer is the preset agent that watches your inbox. When a new email arrives, Laureo reads the thread and the linked CRM context and prepares two reply draftswith different angles. When you open the email, the chooser is waiting in the reply box — pick the draft you prefer, edit anything you like, and send when you’re happy. Newsletters and no-reply mail are skipped.
You manage it from the Agents home like any other preset: pause or resume it, and open its run history to see every draft it produced, with cost and outcome. Drafts are only visible to the mailbox owner.
Only let the agent run when the record is worth its cost.
Because agent steps are metered, it is often worth placing a Filter or If / Else before one so the model only runs on records that matter. The agent step itself can also carry a condition the record must meet first, keeping spend focused.