Front and Knowledge: Refund Policy Cited Draft Reply Template

This Spojit template watches your Front shared inbox and, whenever a customer asks about a refund, drafts a policy-cited reply for an agent to confirm and send.

What It Builds

A Webhook trigger fires when a new Front message arrives. A Condition node checks whether the message mentions a refund. If it does, a Knowledge node queries your refund and returns policy collection, then an Agent-mode Connector node drafts a reply grounded in the matched policy and attaches it as a draft on the Front conversation. Nothing sends automatically: a human agent reviews, edits, and sends.

The Prompt

Paste this into Miraxa and it builds the workflow, connecting the tools for you:

Build a workflow that triggers when a new message arrives in Front. If the message mentions a refund or return, query my "Refund and Returns Policy" knowledge collection for the relevant policy, then have an AI agent draft a reply that answers the customer using that policy and cites it. Attach the draft to the Front conversation so an agent can review and send it. Skip messages that are not about refunds.

Connectors Used

Front - the support inbox that triggers the run and receives the draft reply.
Knowledge - a persistent collection holding your refund and returns policy, queried for the matching passage.
Webhook trigger - starts the workflow on each new Front message.

Customize It

Change the collection name to match yours, widen the keywords in the refund check (add "chargeback" or "money back"), or adjust the draft tone in the prompt. To route only high-value cases for AI drafting, add a threshold the Condition node can read from the message.

Tips

Embed your policy docs into the Knowledge collection first so queries return real citations.
Keep the Connector node in Agent mode so it can phrase the reply; use a Response Schema if you want structured draft fields.
Leaving the reply as a draft keeps a human in the loop before anything reaches the customer.