A ready-to-use system prompt for a refund and returns agent that explains your policy precisely from your own content, cites it, and requires human approval before initiating any refund or return. Copy it, fill the brackets, and paste it into your agent's instructions.
What this agent does
Refund questions are sensitive: customers want a clear, fair answer, and a wrong one costs money or goodwill. This prompt makes the agent explain eligibility, windows, and process strictly from your returns policy, with a citation, so the answer is accurate and traceable. Through MCP it can initiate a return or refund, but only behind human write-confirm — so a person approves before any money moves or any record changes.
The prompt
You are the returns assistant for [Store]. Explain refund and return
eligibility, windows, and process only from the provided context drawn
from [Store]'s policy, and cite the source. If a case is ambiguous or
outside policy, say so and offer to escalate. Never promise a refund the
policy doesn't allow. Any refund or return action requires explicit
confirmation before you proceed.
How to use it
Replace [Store] with your store name, index your returns and refund policy, connect your order system over MCP, and paste the prompt into your agent's instructions. Actions pause for approval automatically. See the e-commerce use case and the billing support prompt, then start free.
Why grounding beats a bare prompt
A prompt alone doesn't make a returns agent reliable — grounding does. Pasted into a generic chatbot, these instructions still let the model answer from its general knowledge and invent details. In AIML.chat the same prompt sits on a retrieval layer that feeds the agent only the content you indexed, so "answer only from the provided context" is enforced by the system rather than merely requested. Every claim links back to a page you control, and the analytics record what people ask so you can refine both the prompt and the content behind it. That pairing — your instructions plus your own grounded content — is what turns this template into a dependable agent.
Will it issue refunds on its own?+
No — it explains policy and can draft an action, but any refund or return requires human approval first.
Can it promise an out-of-policy refund?+
No — the prompt instructs it to answer only from your policy and escalate ambiguous cases.
Is there a free plan?+
Yes — 100 messages a month and one website on the free tier. See pricing for higher limits.