31 lines
2.2 KiB
JSON
Executable File
31 lines
2.2 KiB
JSON
Executable File
{
|
|
"id": "n8n-ai-workflow-automation",
|
|
"name": "n8n AI Workflow Automation Hub",
|
|
"description": "Visual AI workflow automation with n8n at the center. Connect local Ollama models to webhooks, email, databases, and scheduling. Build AI-powered automations without writing code — the exact setup from the viral 'Automate Everything with Local AI' videos.",
|
|
"version": "1.0",
|
|
"author": "ai-lsc",
|
|
"tags": ["n8n", "workflow", "automation", "no-code", "youtube-trending", "local-first", "integration"],
|
|
"endpoints": {
|
|
"n8n": "http://localhost:5678",
|
|
"ollama_base": "http://localhost:11434/v1",
|
|
"redis": "localhost:6379",
|
|
"postgresql": "localhost:5432"
|
|
},
|
|
"tools": [
|
|
"ollama",
|
|
"litellm",
|
|
"n8n",
|
|
"redis",
|
|
"postgresql",
|
|
"fabric",
|
|
"whisper"
|
|
],
|
|
"notes": {
|
|
"youtube_context": "n8n + local AI is the most viewed AI automation content on YouTube. Channels like @n8n_io (official), @techwithtim, and @lainzworld show workflows like: auto-summarize emails with local LLM, transcribe meetings with Whisper, classify support tickets with Ollama, and generate reports on schedule.",
|
|
"recommended_models": "llama3.1:8b (classification, fast), qwen2.5:14b (summarization), mistral-nemo:12b (general), phi-4:14b (structured output), nomic-embed-text (semantic search)",
|
|
"setup": "n8n runs at localhost:5678 with PostgreSQL for persistence and Redis for queue management. Add the Ollama node (built-in) pointing at localhost:11434. LiteLLM at 4000 provides fallback model routing. n8n's AI Agent node chains multiple LLM calls together in a visual flow.",
|
|
"workflow": "Trigger (webhook/cron/email) → n8n AI Agent → Ollama (localhost:11434) → process result → action (email/slack/database write). Whisper node for audio. Fabric node for text transforms. Multiple agents can collaborate in a single workflow.",
|
|
"tips": "Use n8n's sub-workflow feature to reuse AI processing across multiple automations. Set Ollama keep_alive=5m in n8n config to avoid cold starts between workflow triggers. The AI Agent node's memory feature persists conversation context across workflow runs using Redis."
|
|
}
|
|
}
|