Already Have an Account? Sign In ›

Automate iMessage with n8n: the full build

The longest and most complete of the tutorials. It starts from vocabulary — what a node is, what a webhook is, what an API key is, why phone numbers have to be in E.164 — and ends with a published workflow that sends a three-step cadence and hands replies back to an AI agent. If you have never built an automation before, start here.

What you will be able to do

  • Import the published n8n template and wire your own Tuco API key into it
  • Validate a phone number and branch before you spend a send on a junk lead
  • Check iMessage availability, create the lead, then send a plain-text opener
  • Attach a voice note on the follow-up and use Skip If Replied correctly
  • Choose between the SMS fallback and your own check-iMessage branch — not both
  • Point the reply webhook (or ai.draft) at n8n so answers come back into the workflow

Chapters

Timestamps open YouTube at that moment. 13 sections across 42:55.

  1. 0:00

    Vocabulary first

    Workflows, nodes, webhooks, API keys, workspaces, and the E.164 phone format. Skip ahead if these are already familiar.

  2. 3:49

    Import the template

    Download the JSON from the templates page and import it into n8n from the three-dot menu.

  3. 5:53

    The inbound webhook

    POST rather than GET so the body can carry JSON, and the difference between n8n’s test URL and its production URL.

  4. 7:23

    Map the lead fields

    Drag first name, last name and phone out of the webhook output so every run picks up that run’s lead.

  5. 9:07

    Validate the phone number

    Branch on whether the number is real E.164. A junk number exits here instead of burning a send.

  6. 10:53

    Create the API key

    Integrations → API keys in the Tuco dashboard, then bearer auth in n8n. The key is shown once and is scoped to one workspace.

  7. 14:22

    Check iMessage, then create the lead

    The availability check feeds a true/false branch; the true branch creates the lead in Tuco and hands back a lead ID.

  8. 19:39

    Send the first text

    Plain text, with the lead ID from the previous node and an optional from-line ID when the workspace has more than one line.

  9. 24:57

    Apple’s hard rules

    50 new contacts per line per day, no clickable links until they reply, and no attachments on the very first message.

  10. 26:41

    Voice note on the follow-up

    Upload an MP3, confirm the URL loads in incognito, set voiceNote true so it renders as a waveform, and set Skip If Replied.

  11. 32:00

    Fallback or your own branch

    Either let Tuco fall back to SMS through your configured provider, or branch yourself on the check-iMessage node. Pick one.

  12. 32:46

    Receiving replies

    Create the reply webhook in Tuco, point it at n8n, and swap the test URL for the production URL before you publish.

  13. 38:01

    AI agent modes

    Manual listens for the reply event; Copilot listens for ai.draft and needs a website or a custom answer configured. Never both.

Next

Other walkthroughs

All video tutorials · Subscribe on YouTube