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

# Creating an agent

> Walk through the seven-section wizard end-to-end.

This page walks through every section of the agent wizard. If you'd rather just get something on a phone, see the [quickstart](/getting-started/quickstart).

## 1. Start a new agent

From the sidebar, click **Agents → New agent**. Choose:

* **Blank** — start from scratch
* **Template** — start from a preset (sales, support, scheduler, etc.) and customize

Both create a draft you can edit freely.

## 2. Purpose & Personality

The first section seeds the rest of the prompt.

| Field       | What to write                                                      |
| ----------- | ------------------------------------------------------------------ |
| Industry    | The vertical your agent operates in (e.g. *home services*, *SaaS*) |
| Purpose     | One or two sentences — what is this agent for?                     |
| Personality | A few adjectives — *warm, concise, professional*                   |
| Language    | Primary language for the conversation                              |
| Voice       | The TTS voice. Click **Preview** to hear samples                   |

Anyreach uses these inputs to auto-generate a starting prompt in the next section.

## 3. Agent Instructions

This is the heart of the agent — the conversation flow.

Click **Generate instructions** to draft a flow from your purpose, then edit. A good flow includes:

* An **opening line** the agent always says
* The **goals** it should accomplish
* The **questions** it should ask, in order
* A **closing** behavior (confirm next steps, hang up gracefully)

Instructions support markdown. Use bullet lists for steps, and keep imperative voice (*"Ask for the customer's name"* > *"You should ask for the customer's name"*).

## 4. Rules & Boundaries

Hard rules the agent must follow regardless of conversation flow. Examples:

* Never quote pricing — defer to a human
* Never make medical or legal recommendations
* If asked about competitors, redirect politely

These are appended to the prompt with strong emphasis. Keep them short and unambiguous.

## 5. Knowledge

Click **Attach knowledge base** and pick one or more KBs. The agent will retrieve from these whenever a caller asks a factual question.

If you haven't built a KB yet, see [Uploading documents](/knowledge-bases/uploading-documents).

## 6. Abilities

Tools the agent can call mid-conversation. See [Tools](/agents/tools) for the full reference. Common picks:

* **Transfer to a human** — cold or warm handoff
* **Workflow** — invoke a published workflow with parameters
* **Knowledge base lookup** — explicit query (the agent already does implicit retrieval, but explicit lookups are useful when you want a specific KB queried for a specific reason)
* **DTMF** — send tones to navigate IVR menus

## 7. Advanced Instructions

Optional. Lets you override specific prompt sections (e.g. silence handling, interruption behavior, how to handle profanity). Defaults are good for most agents — skip this section unless you're tuning.

## 8. Deployment

Where you go live.

<Steps>
  <Step title="Assign phone numbers">
    Pick from your provisioned numbers, or buy a new one.
  </Step>

  <Step title="Publish">
    Click **Publish** in the top-right. This freezes a new version and routes traffic to it.
  </Step>

  <Step title="Optional: deploy as SIP">
    For BYOC setups, click **Deploy as SIP** to expose a SIP endpoint instead of a Twilio number.
  </Step>
</Steps>

## Iteration loop

Most teams iterate as follows:

1. Test in the simulator after every prompt change ([Testing agents](/agents/testing-agents))
2. Place a few real calls from your phone after each publish
3. Review transcripts in **Conversations**, look for failure patterns, refine prompts and rules

Don't try to make the agent perfect on day one — ship something safe, learn from real calls, iterate.