FlowKit

4 n8n workflows · JSON + guide

AI Inbox Pack

An inbox that sorts, prioritizes and summarizes itself

Four n8n workflows that plug an LLM into your mailbox: every incoming email gets classified, urgent ones surface in Slack in real time, a daily digest summarizes the rest, and AI even drafts your replies in Gmail.

€79VAT included · one-time

Buy the AI Inbox Pack79 €Secure payment by Stripe · Apple Pay · Google Pay · One-time payment · 14-day money-back guarantee

You pay → instant access + email → findable at any time (Find my purchases) · Questions: admin@tech-bharat.com

Instant download · Setup guide in French — workflows are language-agnostic

All 3 packs, 12 workflows: Complete FlowKit Bundle — €269 (€347 bought separately).

The pack

What this pack changes

You spend over an hour a day sorting emails, 80% of which don't deserve your immediate attention. This pack installs intelligent triage upstream: a language model reads every message, categorizes it (urgent, client, admin, newsletter), assigns a priority score, and only interrupts your day when it's justified. Everything else waits for you, summarized, in a daily Slack or Telegram digest.

Who it's for

  • Freelancers handling a contact or support inbox on their own
  • Small teams with an overflowing shared mailbox
  • Any SMB that wants a first AI project that pays off within an afternoon

What's inside

  • 4 n8n workflows as JSON files, importable in two clicks
  • Step-by-step setup guide (in French — workflows are language-agnostic)
  • Classification and drafting prompts already written, ready to customize
  • SQL schema for the queue and logging tables

Requirements

  • n8n ≥ 1.60 (cloud or self-hosted)
  • An OpenAI API key (plus Anthropic for the Gmail drafts workflow)
  • A Slack workspace or Telegram bot for alerts and the digest
  • A free Supabase project for the email queue and logging
  • A Gmail account or any standard IMAP mailbox

What you receive, exactly

The pack-inbox-ia.zip archive, file by file

This list is read from the delivered files each time the site is built: what you see here is what you download.

FileSizeNodes
GUIDE-INSTALLATION.md6.7 KB—
workflows/brouillons-reponse-ia-gmail.json6.6 KB5 + 2 notes
workflows/digest-quotidien-emails-slack.json6.2 KB7 + 2 notes
workflows/priorisation-urgence-emails.json10.8 KB9 + 3 notes
workflows/tri-emails-ia-imap.json15.3 KB9 + 3 notes

Setup guide contents (in French)

  1. Contenu du pack
  2. Prérequis
  3. Créer les tables Supabase
  4. Importer les workflows dans n8n
  5. Configurer les credentials, node par node
  6. Variables à personnaliser
  7. Personnaliser les prompts
  8. Dépannage
  9. Support et garantie

1,103 words · 1 SQL script

Known limits

  • This is not a hosted service: the workflows run in your own n8n instance and model usage (OpenAI, Anthropic) is billed to your own account.
  • The reply-draft workflow only works with Gmail (drafts API); triage uses IMAP and accepts any provider.
  • The classification and drafting prompts are written in French: for a mailbox in another language, you need to adapt them.
  • The digest reads emails with status "pending" but does not mark them as processed: unless you add a status-update node (or clean the table), an email shows up again in the following digests.
  • An LLM can get a category or a score wrong: drafts are never sent on their own, they wait for your review in Gmail.
  • The queue and the log are designed for Supabase; using another database means replacing the relevant nodes (not covered by the guide).
  • The setup guide is written in French only.

Each workflow, real diagram

The 4 workflows, in detail

01 · AI email triage

Sort your emails automatically with AI in n8n (IMAP)

This n8n workflow plugs a language model straight into your IMAP mailbox: every incoming email is read, classified into one of four categories and routed accordingly — Slack alert for urgent ones, Supabase logging for everything.

See how it works node by node
Canvas generated from the delivered JSON exportworkflows/tri-emails-ia-imap.json · 9 nodes · n8n
Email entrant (IMAP) — n8n-nodes-base.emailReadImap v2Email entrant (IMAP)Email (IMAP)Préparer l'email — n8n-nodes-base.set v3.4Préparer l'emailSet (fields)Classifier l'email — @n8n/n8n-nodes-langchain.chainLlm v1.5Classifier l'emailLLM chainModèle OpenAI — @n8n/n8n-nodes-langchain.lmChatOpenAi v1.2Modèle OpenAIOpenAI modelFormat de sortie — @n8n/n8n-nodes-langchain.outputParserStructured v1.2Format de sortieStructured outputRouter par catégorie — n8n-nodes-base.switch v3.2Router par catégorieSwitch (routing)Alerte Slack urgent — n8n-nodes-base.slack v2.2Alerte Slack urgentSlackFusion — n8n-nodes-base.noOp v1FusionNo-opJournaliser dans Supabase — n8n-nodes-base.supabase v1Journaliser dans Su…Supabase
data flow AI sub-node3 explanatory sticky notes in the canvas (not drawn)

02 · email prioritization

Detect urgent emails with an AI score in n8n

This workflow exposes a scoring webhook: send it any email (subject, sender, body) and an LLM assigns an urgency score from 1 to 5, justified and paired with a suggested action. Above the threshold, Slack rings; below it, the email quietly waits for the daily digest.

See how it works node by node
Canvas generated from the delivered JSON exportworkflows/priorisation-urgence-emails.json · 9 nodes · n8n
Email reçu (Webhook) — n8n-nodes-base.webhook v2Email reçu (Webhook)WebhookExtraire les champs — n8n-nodes-base.set v3.4Extraire les champsSet (fields)Scorer l'urgence — @n8n/n8n-nodes-langchain.chainLlm v1.5Scorer l'urgenceLLM chainModèle OpenAI — @n8n/n8n-nodes-langchain.lmChatOpenAi v1.2Modèle OpenAIOpenAI modelScore structuré — @n8n/n8n-nodes-langchain.outputParserStructured v1.2Score structuréStructured outputUrgent ? — n8n-nodes-base.if v2.2Urgent ?IF (condition)Alerte Slack — n8n-nodes-base.slack v2.2Alerte SlackSlackMettre en file pour le digest — n8n-nodes-base.supabase v1Mettre en file pour…SupabaseRépondre au webhook — n8n-nodes-base.respondToWebhook v1.1Répondre au webhookWebhook response
data flow AI sub-node3 explanatory sticky notes in the canvas (not drawn)

03 · email digest

Get a daily email digest in Slack with n8n

Natural companion to urgency scoring, this workflow turns the non-urgent email queue into a daily digest: every morning at 8 am, an LLM reads the Supabase queue, ranks, summarizes and posts the result to Slack and Telegram.

See how it works node by node
Canvas generated from the delivered JSON exportworkflows/digest-quotidien-emails-slack.json · 7 nodes · n8n
Tous les jours à 8h — n8n-nodes-base.scheduleTrigger v1.2Tous les jours à 8hSchedule (cron)Emails en attente — n8n-nodes-base.supabase v1Emails en attenteSupabaseAgréger — n8n-nodes-base.aggregate v1AgrégerAggregateRédiger le digest — @n8n/n8n-nodes-langchain.chainLlm v1.5Rédiger le digestLLM chainModèle OpenAI — @n8n/n8n-nodes-langchain.lmChatOpenAi v1.2Modèle OpenAIOpenAI modelEnvoyer sur Slack — n8n-nodes-base.slack v2.2Envoyer sur SlackSlackEnvoyer sur Telegram — n8n-nodes-base.telegram v1.2Envoyer sur TelegramTelegram
data flow AI sub-node2 explanatory sticky notes in the canvas (not drawn)

04 · AI email replies

Generate AI reply drafts in Gmail with n8n

This workflow watches your Gmail, filters out the noise (newsletters, no-reply), then asks Claude to write a reply proposal saved as a draft in the conversation thread. AI prepares, you decide: nothing leaves without your review.

See how it works node by node
Canvas generated from the delivered JSON exportworkflows/brouillons-reponse-ia-gmail.json · 5 nodes · n8n
Nouvel email Gmail — n8n-nodes-base.gmailTrigger v1.2Nouvel email GmailGmail (trigger)Écarter newsletters et no-reply — n8n-nodes-base.filter v2.2Écarter newsletters…FilterRédiger la réponse — @n8n/n8n-nodes-langchain.chainLlm v1.5Rédiger la réponseLLM chainModèle Claude — @n8n/n8n-nodes-langchain.lmChatAnthropic v1.2Modèle ClaudeAnthropic modelCréer le brouillon Gmail — n8n-nodes-base.gmail v2.1Créer le brouillon …Gmail
data flow AI sub-node2 explanatory sticky notes in the canvas (not drawn)

AI Inbox Pack — €79

One-time payment, instant download, updates included. 14-day money-back guarantee: if a workflow doesn't import, we refund you.

Or get the Complete Bundle: all 3 packs for €269 (€347 bought separately).

Buy the AI Inbox Pack79 €Secure payment by Stripe · Apple Pay · Google Pay · One-time payment · 14-day money-back guarantee

You pay → instant access + email → findable at any time (Find my purchases) · Questions: admin@tech-bharat.com

How it works

Three steps, no surprises

  1. You pay on Stripe

    By card, Apple Pay or Google Pay, on Stripe's checkout page: your card details never touch our servers. One-time payment, receipt sent by email.

  2. You download right away

    The ZIP link (JSON workflows + setup guide) appears on the confirmation page, is also emailed to you and stays findable at any time (“Find my purchases”, at the bottom of every page).

  3. Something wrong? We refund you

    For 14 days, full refund on a simple email to admin@tech-bharat.com, no questions asked.

Seller: LACANGEL GROUP (SAS, Paris, France, VAT FR21995268018). Prices VAT included. Terms of sale

Go deeper

Blog guides related to this pack

FAQ

Frequently asked questions

Does it work with mailboxes other than Gmail?

Yes. The triage workflow uses a standard IMAP node: it works with Gmail, Outlook, OVH, Fastmail or any provider that exposes IMAP. Only the reply-drafting workflow is Gmail-specific (it uses the drafts API).

Are my emails sent to a third-party service?

Email content only passes through your n8n instance and the API of the model you choose (OpenAI or Anthropic, with your own key). Nothing goes through FlowKit. Self-hosted, you keep full control of the pipeline.

How much does the AI usage cost?

The workflows default to economical models (gpt-4o-mini). For an inbox receiving 100 emails a day, expect a few cents to €1–2 per month depending on message length.

Can I change the triage categories?

Yes, it's designed for that: categories live in the classification prompt and the Switch node. The guide shows exactly what to edit to go from 4 to 6 categories or adapt the labels to your business.