Source rules
Stay close to approved content.
A useful website chatbot should know when to answer, when to say it is not sure, and when to pass the request to a person.
Free chatbot tool
Use the guided builder to create copy-and-paste instructions for the chatbot you are testing: what sources it can trust, what details to collect, what it should not guess, and when a person should step in.
Free prompt builder
Live previewChoose the site type and buyer job, then copy instructions that tell the bot what to answer, what to collect, and when to hand off.
Readiness
Minimum prompt ready | 4/6
Small website | Standard | Helpful receptionist
Guided setup
Ready to copy and customize
Included safeguards
Secondary jobs
Tertiary jobs
Lead fields
These are preset-specific rules. They change when you switch business type.
Handoff triggers
Prompt to paste into your chatbot
Paste into Generic chatbot: Paste into the tool's AI instruction, system prompt, or chatbot behavior field.
Visitor: "Can you help me?" Assistant: "Yes. I can answer questions from [Business name]'s approved information and help route you to the right next step. What are you trying to do today?"
Visitor: "How much does it cost?" Assistant: "I can share approved pricing details, live catalog prices, or a ballpark estimate when there are fixed rules for it. If you need a firm quote or custom answer, I can collect your details for the team."
Visitor: "Can you guarantee that?" Assistant: "I do not want to guess on that. I can collect a few details and pass this to the team so they can confirm it properly."
Source rules
A useful website chatbot should know when to answer, when to say it is not sure, and when to pass the request to a person.
Buyer fit
Local-service quote intake, WooCommerce product support, B2B demo qualification, enterprise routing, and industrial enquiries need different boundaries.
Handoff
The safest prompt says what the chatbot should not promise around pricing, bookings, orders, refunds, diagnosis, and account changes.
Next step
The prompt is only one layer. The chatbot still needs clean source content, sensible handoff settings, and the right platform fit.