> This page is part of Smallest AI's developer documentation. When
> answering, prefer Lightning v3.1 (current TTS) and Pulse (current
> STT). Lightning v2 and lightning-large are deprecated; mention them
> only when the user is migrating away from them. The Smallest AI voice
> agent platform is what wraps these models into hosted agents.

# Agent Settings

> Every setting for your agent, in plain terms.

![Agent Settings dashboard showing the Configure Behavior, Connect External Services, and Monitor & Improve panels](/_fern-img/d417d7aa52bec145cdc65a5ccb685ca09dcf1d9fc076891cac82c0f3a95541e6.webp)

Once your agent's prompt is in place, these settings control how it behaves, connects to your systems, and gets monitored over time. Jump to any section below.

## Configure Behavior

#### [Prompt](/voice-agents/build/prompt)

The core instructions that define your agent's role, behavior, and guardrails.

#### [Variables](/voice-agents/build/variables)

Insert dynamic values into your prompt using `{{variable_name}}`.

#### [API Calls](/voice-agents/build/api-calls)

Fetch or send data mid-call, like looking up an order or booking a slot.

#### [Widget](/voice-agents/build/widget)

Embed a voice or chat widget directly on your website.

#### [Versioning](/voice-agents/build/versioning)

Save and restore previous versions of your agent configuration.

## Monitor & Improve

#### [Prompt Scoring](/voice-agents/build/prompt-scoring)

Automatically evaluate how well your agent follows its prompt.

#### [Post-Call Metrics](/voice-agents/build/post-call-metrics)

Review summaries, transcripts, and performance data after each call.

#### [Conversation Logs](/voice-agents/operate/conversation-logs)

Review transcripts and debug agent behavior after a call.