Luna AI - n8n

Meet Luna AI - an Advanced Multi-Agent n8n workflow that uses a Central AI Router to classify user queries and delegate them to six specialized domain agents.

It seamlessly integrates tools like Google Gemini, SerpAPI, Live RSS Feeds, YouTube Analysis & VoiceRSS TTS for intelligent Research, Coding, Writing & Media Generation.



:gear: Agent Components

# AGENTs DESCRIPTION AI MODELs
:one: Router Agent Analyzes user input and intelligently routes queries to the appropriate specialized agent Gemini 3.1 Flash Lite
:two: General Agent Handles everyday conversations, weather, calculations, and general assistance Gemini 3.1 Flash Lite
:three: Research Agent Performs in-depth web research, Wikipedia lookups, YouTube analysis, and multi-modal tasks Gemini 3.1 Flash Lite
:four: Coding Agent Provides coding assistance, debugging, GitHub integration, and multi-language code generation Gemini 3.1 Flash Lite
:five: Writing Agent Creates content, handles translations (EN/BN/HI), and produces creative writing Gemini 3.1 Flash Lite
:six: Media Agent Generates images, videos, and text-to-speech, plus classifies media generation requests Gemini 3.1 Flash Image (Nano Banana 2), Veo 3.0, VoiceRSS API
:seven: News Agent Synthesizes real-time news, RSS feeds (Republic, NDTV, ABP), breaking stories, and web search Gemini 3.1 Flash Lite
1 Like

This is a well-structured multi-agent setup, the router-plus-specialized-agents pattern scales a lot better than trying to cram everything into one prompt. Splitting news, media, coding, and research into separate agents with their own tool access also makes debugging much easier when something goes wrong, since you can isolate which agent misfired instead of untangling one giant prompt. Nice work integrating that many tool sources (RSS, YouTube, SerpAPI, TTS) into a coherent routing layer.

1 Like