Adding Agents to a Scenario

An agent is a model element brought to life inside a scenario. Each agent has a type (one of the node types you put in the twin's scope) and tracks its own values over time. You add agents to a scenario by placing them on a simulation diagram — the same drag-and-drop you already know from regular modeling.

Open or create a simulation diagram

In the Tree View, find your scenario and expand its Diagrams branch
Use the + button to create a new simulation diagram, or open an existing one

The simulation diagram is a canvas just like the ones you use for regular modeling — but tied to one scenario.

Drag agents onto the canvas

Look at the Simulation Toolbox on the left — it shows all the M1 nodes in your model, grouped by type as collapsible accordions
Find the node you want to bring to life (e.g., "Engineering Team", "Alice")
Drag it onto the canvas

Each drag creates one agent bound to that M1 node. The agent inherits the node's type, properties, and any baseline values you've already set.

One node, many agents

For some patterns you want many agents of the same type — five hundred order agents, ten warehouse agents per region, one batch agent per production run. Drag the same M1 node onto the canvas multiple times: each drop creates an independent agent. They share the type's schema (and any formulas) but track their own values.

This is the "start simple, scale up" pattern in action: the metamodel and node only need to be defined once; the simulation can spin up as many agents as it needs.

Linking agents

Once agents are on the canvas, draw connections between them just like in regular modeling — hover over an agent to find the connection handles, drag from source to target, pick a relationship type. These are links — the simulation-level counterpart of the relationships in your model. Agents read each other's state through links.

If two agents both have a corresponding M1 relationship between their underlying nodes, the link can be inherited automatically.

Removing agents

Delete on the canvas removes only the placement — the agent itself stays in the scenario, ready to be placed elsewhere
Delete from the Tree View (or from the agent properties panel) removes the agent from the scenario for good

This mirrors how modeling diagrams work — placement and existence are separate concerns.

Next steps

Entering Time Series Data — give your agents values over time
Simulation Diagrams — work with multiple diagrams per scenario
Writing Property Formulas — let agents compute their own values