Skip to content

README.md

Latest commit

聽

History

History
80 lines (61 loc) 路 4.71 KB

README.md

File metadata and controls

80 lines (61 loc) 路 4.71 KB
Raw

Production-Ready n8n Workflows

This repository contains a collection of production-ready n8n workflows, cleaned and configured for easy deployment. These workflows utilize various AI agents and automation tools.

馃搨 Included Workflows

  1. AI Email Automation.json: Intelligent email auto-responder using RAG (Qdrant) and LLMs.
  2. AI Chat Agent.json: A general-purpose AI chat agent with web search capabilities (SerpAPI).
  3. Instagram DM Agent.json: Automates Instagram DMs using OpenAI and ManyChat integration.
  4. WhatsApp Chatbot.json: A RAG-based WhatsApp sales agent (Yamaha speakers demo).
  5. Podcast Summarizer.json: Summarizes transcripts and sends a formatted email digest.
  6. Ollama Weather.json: Local AI agent (Llama 3.2) with weather lookup capabilities.
  7. Simple Agent.json: A basic Google Gemini (PaLM) powered chat agent.
  8. Reddit AI.json: Monitors Reddit for keywords and uses AI to classify/summarize posts.
  9. Email Scheduler.json: A utility workflow for scheduling emails.

馃殌 Free Deployment Guide

Since many of these workflows use AI Agents, they require a persistent n8n instance with decent memory (~1GB). The most robust "Free" method is Local Hosting + Tunneling.

The "Local + Tunnel" Stack (Recommended)

This method runs n8n on your own computer (Mac/Windows/Linux) but makes it accessible to the internet (for Webhooks, Instagram, WhatsApp) via a free Cloudflare Tunnel.

Step 1: Install & Run n8n (Using Docker Compose) This is the most reliable method. It creates a container that runs exactly like a production server.

  1. Ensure you have Docker Desktop installed.
  2. Open a terminal in this folder and run: 
    docker-compose up -d
  3. Open your browser to http://localhost:5678.

To stop the server later, run docker-compose down.

Step 2: Expose to Internet (Tunnel) To let Instagram/WhatsApp send messages to your local n8n, you need a public URL.

  1. Install Cloudflared (or use ngrok if you prefer).
    • Mac: brew install cloudflare/cloudflared
  2. Start the tunnel: 
    cloudflared tunnel --url http://localhost:5678
    • Or for ngrok: ngrok http 5678
  3. Copy the URL (e.g., https://dry-pond-1234.trycloudflare.com). This is your Webhook Domain.

Alternative: Railway (Free Trial)

Railway offers a free trial that can run n8n, but you will need to add a credit card eventually.

  1. Go to railway.app.
  2. Click "New Project" -> "Deploy a Template".
  3. Search for "n8n".
  4. Deploy.

馃攽 Configuration Matrix

Each workflow requires specific accounts and API keys. Use this table to configure them after importing.

Workflow File Required Services specific Setup Notes
Simple Agent.json Google Gemini Create a "Google Gemini(PaLM) Api" credential in n8n.
AI Chat Agent.json OpenAI, SerpAPI Requires an OpenAI API Key and a SerpAPI Key for web search.
Instagram DM Agent.json OpenAI, Instagram 1. Create a Facebook Developer App.
2. Add "Instagram Graph API".
3. Connect a Facebook Page & Instagram Business Account.
4. In n8n, create "Facebook Graph API" creds.
5. Copy n8n Webhook URL to FB App Dashboard.
WhatsApp Chatbot.json OpenAI, WhatsApp 1. Create a Meta App with "WhatsApp" product.
2. Get a temporary (or permanent) Access Token.
3. Verify a phone number.
4. Copy YOUR_PHONE_NUMBER_ID into the workflow nodes.
AI Email Automation.json IMAP, SMTP, OpenAI, Qdrant Complex Setup.
1. Needs a running Qdrant instance (use Qdrant Cloud free tier).
2. Replace YOUR_QDRANT_URL and YOUR_COLLECTION in the HTTP nodes.
3. Replace YOUR_GOOGLE_DRIVE_FOLDER_ID for file downloads.
Podcast Summarizer.json OpenAI, Gmail Needs Gmail OAuth2 credentials (requires GCP Console setup).
Ollama Weather.json Ollama (Local) 1. Install Ollama.
2. Run ollama run llama3.2.
3. Ensure n8n can reach Ollama (if in Docker, use host.docker.internal).
Reddit AI.json OpenAI Simple setup. Just add OpenAI Key.
Email Scheduler.json Gmail Needs Gmail OAuth2 credentials.

馃洜 Troubleshooting

  • "Webhook not working": Ensure your Tunnel (Cloudflare/ngrok) is running and the URL in the 3rd party app (Instagram/WhatsApp) matches the current tunnel URL.
  • "Workflow inactive": You must click "Activate" (toggle switch) in n8n for triggers to work in the background.
  • "Memory Error": If running AI Agents on free cloud tiers (like Railway starter), they might crash. Local hosting is better for memory-heavy AI tasks.