Operations Center

Operations Center

The Operations Center is where agents work on a project.

• An agent measures the environment's live storefront or the project's content.
• It explains what is off and records what it found.
• Where it proposes a change, a person approves it first.
• Every run is started by a person and paid in credits. Nothing runs on a schedule.

Open Operations Center in the project sidebar, after Studio. It always shows the environment you are in.

Open the Operations Center in the Cockpit

Three agents in every project

Every project has three agents from the start. Nobody adds them, nobody tests them, and having them costs nothing.

They run only when a person starts them, after seeing the estimated credits. More agents can be added: see Agent set-up.

The page at a glance

The page has four tabs: Signals, Findings, Actions and History. Signals shows, top to bottom:

• Banner: until the first measurement, Run your first measurement points to Measure now.
• Measured in your project: real numbers of the environment from the Cockpit itself, such as deployments of the last 30 days, installed apps, markets and languages, and content entries. No agent and no credits are involved.
• Toolbar:
  • Period: 24 hours, 7 days or 30 days.
  • Compare with: the previous period, or no comparison.
  • Source: all agents, or one group.
  • Read again: reloads the signals and keeps your filters.
  • Measure now: runs all agents at once, after one confirmation with every agent's estimate and the total.
• Groups: one per agent, such as Performance, SEO and GEO, and Commerce. A group's head counts its cards by status and links to the agent's findings. Customize chooses which signals you see; the choice applies only to you.
• Cards: one per signal.

What a card shows

• Status:
  • On target: the latest value meets the target.
  • Watch: between the target and the critical line.
  • Critical: past the critical line.
  • A signal without a target has no status.
• Value: the latest value in the period, with its unit.
• Change: against the latest value before the period.
• Chart: the values over the period, with the target as a dashed line.
• Findings: how many the agent reported for it.

Before an agent has measured a signal, its card keeps the same layout and shows no value, only a dash and Not measured, with its target where one is set.

Cards marked EXAMPLE

Some cards show what is coming, not your data:

• the Commerce group, for a conversion agent that is not available yet;
• the SEO agent's Mentions in AI answers and Average position, which no agent measures yet.

These cards show example values and carry the tag EXAMPLE. Only cards without that tag are your data. Request on the Commerce group tells the Laioutr team you want the agent; nothing is charged.

Running agents

• Run analysis runs one agent; Measure now runs all of them.
• The confirmation shows the estimated credits first: the median of the agent's last runs or, before its first run, its run limit, which a run never goes past.
• Only the confirmation starts a run. Plan credits pay first, then purchased credits.
• Measure now lists every agent with its estimate, and says why an agent is skipped: a run already open, a measurement held back, your access, or credits the balance or a limit cannot cover.
• While an agent runs, its group shows Running... and updates by itself.

How a run is paid, and what happens when it cannot be covered: How an agent run is paid.

More on the Operations Center

• Findings and actions: the four tabs, Dismiss, Fix, Approve and Reject.
• Agent set-up: an agent's set-up page, pausing and removing, and adding other agents.
• Who may do what: roles and notifications.