n8n Workflow: Reply to LinkedIn Pros with Reach MCP + OpenAI

Automatically reply to LinkedIn prospects sent via Reach MCP—safely and idempotently

This n8n workflow listens for Reach MCP LinkedIn message.received webhooks, verifies each request with HMAC signature validation, drafts a short reply using OpenAI, and posts it back to Reach using idempotency protection to prevent duplicate sends.

What this workflow does

• Receives a webhook from Reach MCP via a POST endpoint when a LinkedIn event is delivered.
• Reconstructs and validates the payload: rebuilds the raw payload, checks timestamp freshness, and computes an HMAC-SHA256 signature to verify the reach-signature header.
• Filters events so it only continues when:
  • the request is correctly signed,
  • the timestamp is within the freshness window,
  • the event type is message.received, and
  • the message is not from a group chat.
• Drafts a reply with OpenAI by sending the prospect details and last message text to an OpenAI chat model, generating a 2–4 sentence response in the same language.
• Sends the reply to Reach MCP using the event ID as an Idempotency-Key via the Reach MCP LinkedIn Messages API, reducing duplicate outputs when webhooks retry.

Use cases

• Turn incoming LinkedIn messages (delivered through Reach MCP) into fast, on-brand replies without manual copy/paste.
• Automate lead response handling for SaaS teams using n8n + Reach MCP webhooks.
• Ensure secure webhook intake (signature verification + freshness checks) for production message automation.

Technical details

• n8n nodes/tools: if, code, crypto, webhook, http request, plus a sticky note for guidance.
• Integrations: Reach MCP (webhook + LinkedIn Messages API) and OpenAI (chat model for reply drafting).
• Security & reliability: verifies reach-signature using HMAC-SHA256 and posts with an Idempotency-Key to prevent duplicate sends.