Book a build call
Back to Help
Guide

Trust and truth

3 min read

A moving character can feel personal, so the rules must stay clear: it is an AI, it can be wrong, and a person remains in control of access and important actions.

Where an answer comes from

The character is the visible layer, not the brain. Answers come from the selected model, the agent's instructions, connected knowledge, and the current conversation. Good instructions help, but they do not turn every answer into a verified fact.

Permission first

Camera and microphone access use the browser's permission prompt. The agent should explain why access would help before asking. A tool that changes something follows its own access rules and approval boundary.

When it does not know

The right answer is a plain “I don't know,” a clarifying question, or a handoff to a person. Do not treat a fluent answer as proof. Check important medical, legal, financial, or safety information with a qualified person.

Before you publish

Ask what data the agent can see, what leaves the browser, what is retained, who can use each tool, and how a visitor can end the session. Publish only when those answers match what the page says.

Go back to What NetShow Alive is.

Still stuck?

The longer documentation covers models, voice, knowledge, and workflows in depth.

Open the documentation