Prompts & persona
Give the assistant a voice, a system prompt, and clear fallback and refusal behaviour.
This is where you shape how the assistant sounds and behaves — its personality, its instructions, and what it does at the edges.
Persona & voice
Give the assistant a name and a voice tone — warm, formal, playful, or your own custom tone — and add brand phrases it should favour.
System prompt
Write the underlying instructions that guide every reply. This is your main lever over how the assistant reasons and responds.
Fallbacks & refusals
Choose what the assistant does when it can’t help — apologise, ask a clarifying question, or your own behaviour — and set the topics it should refuse, and how it should decline.
Language tone
Decide whether the assistant matches the user’s language, always replies in English, or infers the best language for the conversation.