Sort incoming support mail and prepare review drafts

An n8n flow reads support emails, labels them, and writes reply drafts for a person to check.

Done with n8nยท Intermediate

What they did
A developer put together a five-step n8n automation that watches Gmail for new support messages. For each message, it sends the text to an OpenAI-compatible model, which returns a topic label, an urgency rating, a one-line recap, and a response draft with 3-5 sentences. The workflow then records the output in Google Sheets and shares it in Slack so a human can review it before anything is sent.
How it went
The author said it can be wired up in about 20 minutes and that the drafted replies are solid. Commenters pointed out a few flaws: only one message is handled per polling run, the sender value appears as an object, and the author intentionally kept the final send step manual.
Tips
If you want to use this in production, make the code step loop through every email returned in a poll, pull the sender from from.text, and add JSON parse error handling plus a spam filter. Keep a person in the approval step so the workflow never sends mail on its own.

Try it yourself with n8n

Set up an n8n workflow for [support inbox] that looks for new messages, then for each one identifies the subject area, marks it [low/medium/high] priority, and writes a short summary plus a draft answer based on [FAQ or policy source]. Send the summary and draft to [Slack channel] for human review, and stop before any action that sends, deletes, submits, or changes the email goes ahead. If a message cannot be parsed cleanly, flag it for review instead of inventing details, and end by showing the Slack preview for a test email I can approve.

Fill in the [brackets]. Agents can make mistakes, so review anything before it is sent, paid or deleted.

Read the original on community.n8n.io โ†—

1 more report of this

Get 3 agent uses like this every morning

With the day's top AI news. Free, and one click to unsubscribe.