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

# Creating an agent

> Configure a customer agent's persona, knowledge, actions, and settings, then pair it to a channel.

You build and configure customer agents under **Conversations → Manage agents**. Each agent is a self-contained configuration: a persona, the conventions it follows, the knowledge it can read, the actions it can take, and the channels it answers on.

## The workspace

Open **Manage agents** and create or select an agent. Its configuration is organized into tabs:

| Tab           | What you set                                                                                                                               |
| ------------- | ------------------------------------------------------------------------------------------------------------------------------------------ |
| **Persona**   | The agent's identity and voice—who it is and how it speaks.                                                                                |
| **Behavior**  | Reply mode, the [reply classifier](/customer-agents/operator-workflow#the-reply-classifier), and a [playbook](/customer-agents/playbooks). |
| **Knowledge** | What the agent is allowed to read—knowledge-store access and web search.                                                                   |
| **Actions**   | The app endpoints the agent may call to do things on the customer's behalf.                                                                |
| **Settings**  | Model, reasoning effort, and message splitting.                                                                                            |
| **Test**      | A sandbox conversation to try the agent before pairing it to a live channel.                                                               |

## Persona

The persona is the agent's system prompt—its identity, tone, and the high-level instructions it always follows. Write it the way you'd brief a new teammate: who they are, who they're talking to, and what matters.

Keep the persona about **identity and voice**. Specific, repeatable rules ("always greet by first name", "never quote prices on LINE") belong in [guidance cards](/customer-agents/guidance-cards), which keep the persona short and let you scope each rule to the right channel.

## Knowledge

An agent only knows what you grant it:

* **Knowledge store** — the agent reads from your org's knowledge store within the bounds of its assigned **role**. The role determines which folders it can see, so two agents can draw on different knowledge.
* **Web search** — enable it when the agent needs current information it can't get from your own knowledge.

## Actions

Actions let an agent *do* things, not just answer—look up an order, create a ticket, check availability. An action is an endpoint on one of your apps that the agent is allowed to call. You grant actions two ways:

* **Specific endpoints** — pick individual endpoints the agent may call.
* **Whole apps** — grant an app, and the agent can call any of its agent-callable endpoints (including ones you add later).

Only endpoints that opt in to agent calls are available, and access is re-checked on every run—so revoking it on the app side takes effect immediately. See [Cross-App Communication](/cross-app/agent-calling) for how endpoints opt in.

## Settings

* **Model** — leave on the platform default, or pin a specific model.
* **Reasoning effort** — how much the model deliberates before answering.
* **Message splitting** — when on (the default), a reply can be split into separate chat bubbles by putting `---` on its own line.

## Reply mode

Set the agent's default [reply mode](/customer-agents/overview#reply-modes)—**auto-send**, **auto-draft**, or **off**—under **Behavior**. This is the default for new conversations; your team can override it per conversation at any time.

<Tip>
  Start a new agent in **auto-draft** so your team reviews every reply while you tune the persona and
  guidance. Switch to **auto-send** once you trust it.
</Tip>

## Pair it to a channel

A configured agent doesn't answer anything until a channel routes to it. Open **Conversations → Channels** and pair each connected channel to an agent:

* **LINE** and **Facebook** accounts pair here—inbound on that connection routes to the agent you choose. (Instagram DMs come through the Facebook Page connection.)
* **Web** needs no pairing—your app names the agent when it calls the SDK.

Connect the LINE or Facebook accounts themselves first under [Settings → Integrations](https://www.stardeck.ai/dashboard/settings?tab=integrations). See [LINE](/connected-services/line/overview) and the Connected Services overview for connecting accounts.

## Test before you go live

Use the **Test** tab to hold a sandbox conversation with the agent—same persona, guidance, knowledge, and actions, but nothing reaches a real customer. Test conversations are kept separate from the live pool.

## Next steps

<CardGroup cols={2}>
  <Card title="Guidance cards" icon="list-check" href="/customer-agents/guidance-cards">
    Reusable, channel-scoped conventions
  </Card>

  <Card title="Playbooks" icon="route" href="/customer-agents/playbooks">
    Give the agent a goal and a handoff
  </Card>

  <Card title="Operator workflow" icon="inbox" href="/customer-agents/operator-workflow">
    Review drafts and take over conversations
  </Card>

  <Card title="Actions & cross-app calls" icon="bolt" href="/cross-app/agent-calling">
    Let endpoints accept agent calls
  </Card>
</CardGroup>