🛠️ How-to guide

<aside> ⚙️

WHAT IT DOES

The Settings tab is where you set an agent's name, how it converses, which features are on, its privacy and compliance rules, and the AI model it runs on. Part of Inside Your Agent.

</aside>

Before you start

Anyone can view Settings; changing them needs a role that can edit agents — Editor, Publisher, or Admin (see Roles & Permissions). Changes save from the bar at the bottom of the page.

The Settings tab — Identity, Conversation, Features, Privacy & Compliance, Model Configuration, and Post-Processing for a single agent.

Identity

• Agent Name — the agent's name. The lock keeps the AI Designer from renaming it as you build; lock it once you're happy with the name.
• Designer Instructions — the instructions the Designer generated (read-only here; edit them back on the Designing Your Agent tab).
• Additional Formatting & Style — your own tone and structure notes, appended after the Designer instructions (e.g. "Keep answers under 3 sentences.").

Conversation

• Timezone — used for dates and scheduling.
• Conversation Expiry — how long before an idle conversation resets (d.hh:mm:ss; blank = never).
• Supported Languages — specific locales (e.g. en-US, fr-FR); leave empty for any.
• Active Modalities — where the agent runs: LiveText, AsyncText, Voice.

Features

• Suggested Replies — show clickable reply suggestions.
• File Uploads — allow users to send files.
• Show active agent name — display which agent is replying, updating as agents hand off (web chat).
• Verbose Tracing — record detailed per-turn diagnostics in the log; helpful while building, noisy in production — turn it off once the agent is stable.

Privacy & Compliance