Transform GitHub JSON to Telegram Bot with n8n Workflow

Transform your GitHub JSON files into intelligent Telegram chatbots with retrieval-augmented generation (RAG) — no vector databases, no Pinecone subscriptions, just pure n8n automation that turns static knowledge bases into conversational AI assistants.

What this workflow does

This n8n automation creates a fully functional Telegram bot that answers questions using data from your GitHub repository. When users send /ask <question> to your bot, the workflow automatically pulls your JSON knowledge base from GitHub, runs intelligent keyword matching to find relevant content, feeds the context to Qwen 3 via OpenRouter, and delivers personalized answers directly in Telegram.

The workflow includes robust validation that rejects messages shorter than 7 characters and guides users with clear instructions: "Please use: /ask <your question>". This prevents unnecessary API calls while ensuring users understand the correct format immediately.

Use cases

• Customer Support: Convert FAQ documents into 24/7 Telegram support bots for instant customer assistance
• Team Knowledge Base: Transform internal documentation into searchable chatbots for quick employee access
• Product Information: Create interactive product catalogs that answer specific feature questions
• Educational Content: Turn training materials into conversational learning assistants
• API Documentation: Make technical documentation more accessible through natural language queries

Technical details

This workflow leverages essential n8n nodes including Telegram Trigger for message handling, GitHub integration for knowledge base retrieval using Personal Access Tokens, Code nodes for keyword matching logic, and conditional IF nodes for input validation. The automation connects seamlessly with OpenRouter's Qwen 3 model for natural language processing.

Built-in error handling ensures users receive specific feedback for invalid GitHub tokens, missing files, or empty LLM responses — no silent failures or generic errors. The local keyword-matching engine eliminates the need for expensive vector database subscriptions while maintaining effective content retrieval.

Perfect for n8n users seeking cost-effective RAG implementations without complex infrastructure dependencies.