Door 02 · For Builders
The n8n workflow powerhouse.
A public collection of diverse, production-shaped n8n workflows focused on AI, agentic systems, and automation. Clone it. Read it. Ship from it.
33
Workflow templates
JSON
Aggregated for LLM analysis
MCP
Tool integration patterns
Public
Clone, read, iterate
01 · Inside the repo
What you're cloning.
{ }
.json workflow files
Ready-to-import templates covering AI document parsing, agentic chatbots, social media automation, and local LLM integration.
⬢
Aggregated raw JSON
One file containing every workflow's JSON — perfect for feeding into a large-context LLM to analyze n8n patterns end-to-end.
≡
Titles + descriptions
A browsable, searchable list (this page) mapping filenames to short descriptions so you can find what you need fast.
02 · Method
Feed it to an LLM.
The aggregated JSON + descriptive metadata aren't just for n8n import — they become a learning substrate when combined with a frontier LLM (Claude, Gemini, GPT — anything with a serious context window). Ask it things like:
- Explain common patterns for integrating local Ollama models across these workflows.
- Walk me through the RAG chatbot workflow node-by-node.
- Draft a new workflow JSON that combines X and Y.
- What are common parameters for the HTTP Request node in these POSTs?
- Generate a README draft from this workflow's JSON.
# Context
I've attached:
- 0. workflows.diy — n8n — public repo.md
(aggregated JSON of all workflows)
- 1. Workflow Titles and Descriptions.md
# Task
Based on the attached workflows, can you explain the
common patterns used for integrating local Ollama models
for private AI tasks? Reference examples like the
"Private & Local Ollama Self-Hosted + Dynamic LLM Router".
# Output
Return a structured breakdown:
1. shared architectural patterns
2. node-level configuration deltas
3. one annotated exampleThis is the AI-First way to learn n8n — pair programming, but for patterns.
03 · Index
33 / 33 workflows.
-
Leverages Llama Parse and AI to parse documents (e.g., from Gmail attachments), extract text/structured data (like invoices), and save results to Google Drive/Sheets.
-
An AI agent chatbot with long-term memory and note-taking capabilities using Google Docs, interacting via Telegram.
-
An AI chatbot equipped with a Jina.ai tool to scrape and answer questions based on real-time web content.
-
Uses an AI agent to query and analyze n8n community leaderboard data (from GitHub) to find and report on popular workflows by specific creators.
-
An AI agent that fetches data from the n8n community leaderboard and generates reports on top creators and their workflows, potentially with scheduled triggers.
-
A Retrieval-Augmented Generation (RAG) chatbot that processes documents from Google Drive, stores them in Qdrant, and uses Google Gemini AI to answer questions based on the documents.
-
Automates the summarization and analysis of YouTube videos by fetching transcripts and using AI to generate insights.
-
Utilizes Google Gemini AI to analyze YouTube videos, providing various outputs like summaries, transcripts, scene descriptions, or highlight clips based on user-selected prompt types.
-
An AI-driven content factory for generating tailored posts for multiple social media platforms (LinkedIn, Instagram, Facebook, X, TikTok, etc.), including image suggestions and an approval loop.
-
A highly advanced social media content factory that uses externally managed system prompts and schemas (from Google Docs) to dynamically generate and publish content across various platforms.
-
Automatically backs up all n8n workflows to a timestamped Google Drive folder and manages older backups.
-
Creates an AI chatbot using GPT-4o that can perform web searches via the MCP Brave Search tool to answer user queries.
-
Analyzes an image using multiple local Ollama vision models (e.g., Granite, Llama3.2-vision) and saves the comparative results to Google Docs.
-
An AI chatbot that fetches content from a specific Confluence page and answers user questions based on that content.
-
An AI agent powered by DeepSeek with long-term memory capabilities (using Google Docs) interacting via Telegram.
-
A quick start guide demonstrating various methods to connect to and use DeepSeek V3 (chat) and R1 (reasoning) models, including local Ollama and direct API calls.
-
Extracts text from PDF documents, uses AI to generate blog posts, creates drafts in WordPress, and includes a Gmail-based human approval step before publishing.
-
An AI chatbot with long-term memory (via Google Docs) and a dynamic tool router that can save, retrieve, or send memories to other services like Gmail or Telegram.
-
Scrapes multiple pages of a website by first fetching its sitemap and then using Jina.ai to extract content from each listed URL.
-
Uses Perplexity API for research on a given topic, then leverages AI to generate SEO-optimized blog content and publish it to WordPress.
-
Monitors specified YouTube channels using their RSS feeds, fetches details of new videos, and sends notifications via Telegram and/or Gmail.
-
A multi-agent chatbot system where a primary agent routes tasks to secondary agents specializing in querying a Postgres/Supabase database or generating charts with QuickCharts.
-
Fetches Nostr threads with the #damus hashtag, uses AI to analyze themes and generate reports, and sends them via Gmail and Telegram.
-
Converts PDF documents into blog posts and publishes them to a Ghost CRM instance using AI for content generation.
-
Performs research on a topic using Perplexity API and then uses AI to generate a structured article, subsequently converting it into a styled HTML document.
-
An AI assistant that uses a router agent to dynamically select the most appropriate locally hosted Ollama LLM (e.g., text, coding, vision models) based on the user's prompt.
-
A simple AI assistant powered by a locally hosted Ollama LLM (e.g., Llama 3.2) for private chat interactions.
-
A Telegram bot that can receive and process text, audio (transcribing with OpenAI), and image (analyzing with OpenAI Vision) messages, with user validation.
-
Utilizes the Tavily API for AI-powered web search and content extraction to perform research on a given topic, optionally summarizing results with AI.
-
An AI agent chatbot that retrieves YouTube video details and transcripts, then allows users to ask questions or request summaries/analyses of the video content.
-
Transcribes audio files (e.g., .m4a from Google Drive) using OpenAI, then uses AI to summarize the transcript into structured JSON and Markdown reports, saving all to Google Drive and notifying via Gmail/Telegram.
-
A quick start guide to create and publish AI-generated blog posts to WordPress, including generating a title, content, and a featured image.
-
An AI agent that fetches YouTube video details and comments, then analyzes the comments to provide a comprehensive report including sentiment, common themes, and content opportunities.
04 · Import
Import a workflow.
-
01Open the
.jsonfile on GitHub. - 02Select all & copy the JSON content.
- 03In n8n, open an empty workflow or the workflows list.
- 04Click empty canvas & paste.
⚠ Configuration is key
-
Most workflows contain
PLACEHOLDERvalues. Replace them with your own keys, URLs, IDs. - Ensure credentials are configured in n8n for services like OpenAI, Google, Telegram.
- Test thoroughly after configuring.
05 · Prereqs
What you'll need.
- An n8n instance (cloud or self-hosted).
- Credentials for services used by a given workflow (OpenAI, Google Cloud, Telegram, WordPress, Ollama API, etc).
-
Optionally, n8n Community Nodes and/or npm packages
(e.g.
youtube-transcript) if a Code node imports externals. - Familiarity with the underlying services helps for troubleshooting and customization.
These workflows are provided as-is, for educational use. Use at your own risk. Be mindful of API costs, rate limits, and each provider's terms of service.
Hire
Want custom n8n + AI built for your stack?
Templates get you 80% there. The last 20% — integration, auth, edge cases, orchestration — is where most projects stall. Hire me to finish the job.