> ## Documentation Index
> Fetch the complete documentation index at: https://docs.erstan.com/llms.txt
> Use this file to discover all available pages before exploring further.

# Create your first agent

> Build, preview, and publish a working agent from scratch in the visual builder.

This page walks you through building your first agent end to end — from picking a shape to publishing a Live agent your team can run. We use the simplest path, an **Assistant**, so you can get a useful agent working in minutes.

<Frame caption="Choosing an agent shape">
  <img src="https://mintcdn.com/erstan/VvzUuagDdyVkLVu5/images/new-agent-create-dialog.png?fit=max&auto=format&n=VvzUuagDdyVkLVu5&q=85&s=4519374cf25cd5b5d7249f6c7264111c" alt="The Create agent dialog showing AI Assistant, Workforce, and Multi-Step Automation shapes" width="2880" height="1800" data-path="images/new-agent-create-dialog.png" />
</Frame>

<Tip>
  Prefer not to build from scratch? Choose **Explore the marketplace** in the **Create a new agent** dialog to install a ready-made, NetSuite-grounded agent instead. See [Adding agents from the Marketplace](/marketplace/add-agents).
</Tip>

## Before you start

<Info>
  Agents need an AI model to think and a connector to act. Make sure your workspace has a connected [LLM provider](/connectors/llm-providers) and, for anything that touches your data, a connected [NetSuite](/connectors/netsuite) account.
</Info>

## Build an Assistant step by step

An **Assistant** is a chat-first agent — a domain expert that plans, retrieves, and acts, looping you in when it needs a decision. It is created with a **Chat Trigger** and a default graph of **Chat Trigger → AI Respond**, so you have a working agent the moment you create it.

<Steps>
  <Step title="Open Agents and start a new agent">
    Go to **Agents** in the workspace, then click **New agent**. The **Create a new agent** dialog opens with the three shapes.
  </Step>

  <Step title="Pick the Assistant shape">
    Choose **Assistant Agent** ("Domain expert"). Erstan creates it immediately with a default name (for example, *New Assistant*) and drops you into the visual builder.

    <Note>
      **Workforce** ("Orchestrator") and **Automation** ("Set process") are the other two shapes. A Workforce routes work across other agents; an Automation runs a multi-step process from a **Form** or **Schedule**. See [Agents overview](/agents/overview) for when to use each.
    </Note>
  </Step>

  <Step title="Name and describe it">
    In the builder, give the agent a clear name and short description (for example, *AR Aging Assistant — summarizes overdue invoices and customer balances*). This is what teammates see in the Agents library.
  </Step>

  <Step title="Confirm the trigger">
    Your Assistant already has a **Chat Trigger**, which means people start it by chatting or `@`-mentioning it. You don't need to add another trigger for a chat-first agent.

    <Note>
      Want it to run from a form or on a schedule instead? That's a different shape — see [Triggers](/agents/triggers).
    </Note>
  </Step>

  <Step title="Add steps (optional)">
    The default **AI Respond** step is enough for a capable assistant. To add behavior, open the node library on the canvas and drag in steps from **Triggers**, **AI Steps**, **By Connector**, or **Utility**, then connect them with edges. See [The visual builder](/agents/builder) and the [Step reference](/agents/step-reference).
  </Step>

  <Step title="Set instructions, skills, and tools">
    Use the inspector to tell the agent how to behave:

    * **Instructions** — plain-language guidance ("Always show invoice number, customer, amount, and days overdue").
    * **Skills** — reusable instruction packages it should load. See [Giving agents tools, skills & connections](/agents/tools-skills-connections).
    * **Tools / connections** — the connector actions it may use, such as NetSuite reads. A step that needs an unconnected system shows a **Connection required** banner until you connect it.

    <Warning>
      Tools that write to NetSuite (creating or updating records) require approval by default. Keep write-capable tools set to **Require approval** unless you have a strong reason to change it. See [Approving and rejecting AI actions](/chat/approvals).
    </Warning>
  </Step>

  <Step title="Preview before publishing">
    Click **Preview** in the builder header to run the agent without making it Live. Provide a chat message, watch the run stream live in the **Steps** panel, then iterate. Preview runs are tagged **Preview run** in run history.
  </Step>

  <Step title="Publish">
    When it works, click **Publish**. The agent's status becomes **Live** and it appears in the library for your team to run.

    <Note>
      For triggered agents (Form, Schedule, Webhook, Email), publishing is gated until the trigger's **Run As** identity is configured and consented. A plain chat Assistant has nothing extra to configure. See [Preview, publish & versions](/agents/publish-lifecycle).
    </Note>
  </Step>
</Steps>

<Frame caption="The visual agent builder">
  <img src="https://mintcdn.com/erstan/VvzUuagDdyVkLVu5/images/agent-builder-canvas.png?fit=max&auto=format&n=VvzUuagDdyVkLVu5&q=85&s=b45c51d1e0dd0976f5cbb9528e1ce4eb" alt="The visual agent builder showing a Chat Trigger connected to an Erstan action node, with the Add Step panel" width="2880" height="1800" data-path="images/agent-builder-canvas.png" />
</Frame>

## Let the AI builder do the work

Inside the builder you can describe what you want in plain language and have Erstan edit the agent for you — adding steps, wiring connections, and writing instructions. Use it to scaffold a first draft, then refine by hand. See [The visual builder](/agents/builder) for details.

## After you publish

A Live agent can be run on demand, assigned as a task, or started by `@`-mention in chat. Every run produces a tracked trace with live status and approvals.

<Check>
  You now have a Live Assistant. Run it from the library or the agent's runs page, and watch its trace in [Run history](/agents/run-history-traces).
</Check>

## Next steps

<CardGroup cols={2}>
  <Card title="The visual builder" icon="workflow" href="/agents/builder">
    Master the canvas, the node library, and the inspector.
  </Card>

  <Card title="Running agents" icon="play" href="/agents/run-and-tasks">
    Run on demand, assign as a task, or schedule recurring runs.
  </Card>

  <Card title="Giving agents tools & skills" icon="wrench" href="/agents/tools-skills-connections">
    Attach the skills, tools, and connectors your agent needs.
  </Card>

  <Card title="Explore the Marketplace" icon="store" href="/marketplace/add-agents">
    Install a ready-made agent instead of building one.
  </Card>
</CardGroup>