Mindket

Gmail draft replies written by GPT for mail that needs one

Newversion 1 · updated

Overview

An n8n workflow that checks each new Gmail message with GPT, skips marketing mail, and saves a drafted reply in the same conversation when an answer is needed. Nothing is sent until you read and send it.

Gmail draft replies written by GPT for mail that needs one is an n8n workflow in the Productivity category, listed by Mindket. It is free. This is version 1, last updated on . Download it on this page or copy its JSON, then import it into an n8n canvas. Nothing runs until you activate it in n8n.

At a glance

Runs in
  • n8n Cloud: import the file and publish it
  • Your own n8n server: the same file and steps
  • Gmail: the drafts appear in the Drafts folder and in the thread

Platform

Connects 2 apps

How it works

Trigger

New email

It starts when a new email that did not come from you reaches your Gmail, checked every minute. It brings:

  • The sender, subject and body
  • The conversation it belongs to

Actions

  1. Asks a first model whether the message needs a reply, and skips marketing mail
  2. Lets only the messages that need an answer go on
  3. Asks a second model to draft a business casual reply in the sender’s language
  4. Saves the reply as a Gmail draft in the same conversation, addressed to the sender

Result

A drafted reply waiting in your Gmail Drafts folder for each message that needs one, ready for you to edit and send

Integrations

What you need before you start.

Gmail

Mail source and drafts

Fetches new messages and saves the drafted replies in their conversations.

NeedsA Gmail OAuth2 credential

OpenAI

Language model

Judges each message and writes the reply. The copy asks for gpt-6-luna and gpt-5.6-sol.

NeedsAn OpenAI API key

Workflow check

Checked by Mindket on 13 Nov 2026

Steps
5 steps, starting on a Gmail event
AI
2 steps call an AI model: Assess if message needs a reply and Generate email reply
Sign-ins
Signs in to: Gmail and OpenAI
API keys
No API key is left in the file
Step types
Mindket’s checker reads all 4 of its step types

Read from the workflow’s file, not from a run.

About

What it is

This n8n workflow writes draft replies in your Gmail. Every minute it fetches new messages that did not come from you. A first model reads each one and answers in JSON with true or false: does it need a reply? Marketing mail does not. For the mail that does, a second model drafts an answer, and Gmail saves it as a draft in the same conversation, addressed to the sender. Nothing is sent. Eight nodes run it, beside four sticky notes with short titles.

Who it suits

Busy people who answer the same kinds of email all day and would rather edit a draft than start from nothing. Skip it if your mail holds private matters you do not want sent to OpenAI.

Who uses it, and how
IndustryWho uses itHow they use itExample
Small businessOwnersGet a first answer ready for each customer emailAn owner opens Drafts and finds a reply to a quote request.
ConsultingIndependent consultantsKeep up with client questions between meetingsA consultant edits a drafted yes-and-no pair for a scheduling request.
SupportShared inboxesPrepare replies for a human to checkA support lead reviews the drafts each morning and sends the good ones.
RecruitingHiring managersAnswer candidates fasterA manager reads a drafted reply to a candidate’s question and sends it.

You need to know n8n’s settings: import a file, add two credentials and publish it.

Where to use it

Any n8n that can reach Gmail and OpenAI will run it.

  • n8n Cloud: import the file and publish it
  • Your own n8n server: the same file and steps
  • Gmail: the drafts appear in the Drafts folder and in the thread

Setup

Download saves a zip with workflow.json and Mindket’s setup notes in your downloads folder. Setup is two credentials and a read of the reply rules.

What you need first
NeedDetail
n8nOn n8n Cloud or on a server of your own
GmailAn OAuth2 credential for the account whose mail you want answered
OpenAIAn API key. The check uses gpt-6-luna and the reply uses gpt-5.6-sol
  1. In n8n, create a workflow, open its menu, choose Import from File and pick workflow.json.
  2. Add a Gmail credential and select it in Gmail Trigger and in Create Gmail Draft.
  3. Add an OpenAI credential and select it in OpenAI Chat and in OpenAI Chat Model.
  4. Open Generate email reply and change its instructions to match your tone, greeting and sign-off.
  5. Publish the workflow.
  6. Send yourself a question from another address and open your Drafts folder.

How to use it

Leave it running. When a message that needs an answer arrives, a draft appears in its conversation. Read it, change it and send it yourself.

What to change for a different result
To changeWhere
Which mail is looked atThe search -from:me in Gmail Trigger
What needs a replyThe message in Assess if message needs a reply
Tone, language and rulesThe message in Generate email reply
The modelsOpenAI Chat and OpenAI Chat Model

Example

A customer writes this to you.

Can you deliver on Friday?

The draft opens with “Hello,” and holds two replies, one that says yes and one that says no, split by a line of dashes. You keep the one that fits and delete the other.

For newcomers and experts

If you are new. n8n draws a workflow as boxes joined by lines, and each box, a node, does one job. This one asks a model whether mail needs an answer, asks a second model to write it and saves a draft. Send a few test emails before you leave it on.

For experts. One chain makes a JSON decision and a second chain writes the text.

Where to look closer
NodeWhat it doesWhat to know
Assess if message needs a replyChain with gpt-6-luna and a JSON parserIt returns needsReply as true or false
If Needs ReplyLets only the true cases onIts check is a strict boolean
Generate email replyChain with gpt-5.6-solThe rules cover tone, greeting, sign-off and two answers for yes-or-no questions
Create Gmail DraftSaves an HTML draftThe thread id keeps it in the conversation

Good to know

Every draft needs your eyes, because the model can state things you never said.

  • Models: the original asked for gpt-4o, which OpenAI shuts down on 14 January 2027, and gpt-4-turbo, which it shut down on 23 October 2026. This copy asks for gpt-6-luna and gpt-5.6-sol, the replacements OpenAI names, and leaves out the temperature
  • Nothing is sent: every reply is a draft
  • Cost and data: each new email is one call to OpenAI, and each email that needs a reply is a second one
  • Search: -from:me also matches automatic mail from other services, so narrow it for fewer calls
  • Placeholders: where the model does not know an answer, the draft holds a note such as [YOUR_ANSWER_HERE]
  • Not run: Mindket did not run it against Gmail or OpenAI
  • Credentials: the JSON holds none

README and files

README

Gmail draft replies written by GPT for mail that needs one

An n8n workflow that writes draft replies in your Gmail. Each new email is first judged by a model: does it need an answer?
Marketing mail does not. For the mail that does, a second model writes a reply and Gmail saves it as a draft in the same
conversation. Nothing is sent. You read each draft, change it and send it yourself.

What is inside

  • workflow.json: the workflow as n8n exports it, twelve nodes. Four are sticky notes with short titles. It holds no
    credential.
  • README.md: these notes.

What we changed from the original copy

  • The first model asked for gpt-4o, which OpenAI shuts down on 14 January 2027. OpenAI Chat now asks for
    gpt-6-luna, the replacement OpenAI names. Its temperature of 0 is gone, because the model's page does not list a
    temperature setting.
  • The reply model asked for gpt-4-turbo, which OpenAI shut down on 23 October 2026. OpenAI Chat Model now asks for
    gpt-5.6-sol, the replacement OpenAI names.
  • The node that saves the draft was called Gmail - Create Draft. It is now Create Gmail Draft.
  • The links to the author's Gmail and OpenAI credentials are gone.

Before you start

  • A Gmail account you can connect to n8n with OAuth2.
  • An OpenAI API key.

Set it up

  1. In n8n, create a workflow, open its menu and choose Import from File. Pick workflow.json.
  2. Add a Gmail OAuth2 credential. Select it in Gmail Trigger and in Create Gmail Draft.
  3. Add an OpenAI credential and select it in OpenAI Chat and in OpenAI Chat Model.
  4. Read the instructions in Generate email reply. They set the tone, the greeting and the sign-off, so change them to
    sound like you.
  5. Publish the workflow. Gmail Trigger checks your mailbox every minute.
  6. Send yourself a question from another address and open your Drafts folder.

How a run goes

  1. Gmail Trigger fetches each new email that does not come from you, using the search -from:me, and parses its
    headers and its text.
  2. Assess if message needs a reply asks OpenAI Chat whether the message needs an answer and expects JSON with one
    true or false value. JSON Parser checks the shape. Marketing emails are told to count as not needing one.
  3. If Needs Reply lets only the true cases through.
  4. Generate email reply asks OpenAI Chat Model for a reply. The instructions ask for a business casual tone, a
    greeting of "Hello," and a sign-off of "Best,". For a yes or no question it writes two replies, one affirmative and one
    negative. Where it does not know an answer, it leaves a placeholder such as [YOUR_ANSWER_HERE].
  5. Create Gmail Draft saves the reply as an HTML draft in the same conversation, with the subject Re: and the
    original subject, addressed to the sender.

Change the result

  • Which mail is looked at: the search in Gmail Trigger. Add terms to skip newsletters or certain senders.
  • What counts as needing a reply: the message in Assess if message needs a reply.
  • The tone, the language and the rules of the reply: the message in Generate email reply.
  • The models: OpenAI Chat and OpenAI Chat Model.

Good to know

  • Nothing is sent. Every reply is a draft.
  • Every new email is one model call, and each email that needs a reply is a second one. All of it goes to OpenAI.
  • The search -from:me also matches automatic mail from other services, so the first model decides what needs an
    answer. Narrow the search if you want fewer calls.
  • The draft is addressed to the email's From header and joins its thread.
  • The reply model may invent details. Read each draft before you send it.
  • Mindket did not run this workflow against Gmail or OpenAI. These notes follow n8n's published node code and OpenAI's
    deprecations page, read on 12 November 2026.
Files
NOTICE.md  673 B
README.md  4 KB
workflow.json  8 KB

3 files, 13 KB, downloaded as a zip.

What’s new

  • Version 1 · 2026-11-12

Reviews

No ratings yet

Write a review

No reviews yet.

More from Mindket

See all from Mindket

anaksor-ocr

Sokheng · Kit

Reads Khmer and English text from documents, receipts and photos of signs, using a detector for text lines and a recogniser for each line. Suits developers who…

New

Qwen2.5-7B-Instruct as GGUF files in eight quantization sizes, to chat with locally through llama.cpp. For people who want a capable 7.6B model on their own co…

New

ACE-Step 1.5 XL SFT

ACE-Step · Kit

Generates music from text with a 4B diffusion decoder, 50 steps and guidance control, aiming for higher audio quality than the base. For creators with a GPU of…

New

Similar in Productivity

See all in Productivity

Guizang PPT Skill

op7418 · Skill

Skill that makes single-file HTML slide decks, slide illustrations and platform covers in two visual styles, with a presenter mode for rehearsal. For speakers…

New29.8k

Frontend Slides

zarazhangrui · Skill

Builds HTML presentations from scratch or from a PowerPoint file and shows visual style previews to pick from, for people who want web slides with no CSS or Ja…

New32.3k

IPTV

iptv-org · Kit

Collects publicly available live TV channels from many countries into playlists that any video player with live streaming can open. Made for viewers and for de…

New144k