Create and publish AI social posts from Telegram with OpenAI, AtlasCloud and Blotato
Needs changesWhat this workflow does
This AI workflow has 17 nodes and runs on Telegram events (message). Along the way it connects Telegram, OpenAI Chat Model, HTTP Request and Blotato (community). It also calls api.atlascloud.ai over HTTP.
How it runs
Trigger: On Telegram events (message).
- Workflow settings (Edit Fields (Set)): sets 6 fields: language, brand_voice, image_style…
- AI content agent (AI Agent): runs an AI agent, with model: OpenAI Chat Model; 1 tool (Read a web link); memory: Conversation memory; output parser: Format the content
- Generate image (Grok) (HTTP Request): calls api.atlascloud.ai (POST)
- Wait: pauses before continuing
- Poll image result (HTTP Request): calls an HTTP API
- Check image status (Switch): routes items into 2 branches
- Extract image URL (Edit Fields (Set)): sets 1 field: image_url
- Stop and Error: stops the run with an error
- Publish to LinkedIn (Blotato): uses Blotato
- Publish to X (Blotato): uses Blotato
…plus 2 more steps. See the full node list below.
Services it uses
HTTP calls to: api.atlascloud.ai
Before you upgrade to n8n 3.0
1 thing in this workflow is removed in n8n 3.0. Fix it before you upgrade:
- HTTP Request Tool (legacy) in Read a web link: replace with HTTP Request node connected to the AI Agent's Tool input.
Rather not rewrite it? Many of these flows can be rebuilt in Make (free plan).
| Removed in 3.0 | Replace with |
|---|---|
| HTTP Request Tool (legacy) Read a web link | HTTP Request node connected to the AI Agent's Tool input |
Nodes used (13)
- Telegram Trigger
n8n-nodes-base.telegramTrigger - Edit Fields (Set)
n8n-nodes-base.set - AI Agent
@n8n/n8n-nodes-langchain.agent - OpenAI Chat Model
@n8n/n8n-nodes-langchain.lmChatOpenAi - Simple Memory
@n8n/n8n-nodes-langchain.memoryBufferWindow - HTTP Request Tool
@n8n/n8n-nodes-langchain.toolHttpRequest - Structured Output Parser
@n8n/n8n-nodes-langchain.outputParserStructured - HTTP Request
n8n-nodes-base.httpRequest - Wait
n8n-nodes-base.wait - Switch
n8n-nodes-base.switch - Blotato
@blotato/n8n-nodes-blotato.blotato - Telegram
n8n-nodes-base.telegram - Stop and Error
n8n-nodes-base.stopAndError
From the author
Quick Overview This workflow takes a Telegram message containing an idea or URL, uses OpenAI to generate platform-ready captions plus an image prompt, creates an image via AtlasCloud (Grok Imagine), then publishes the post to LinkedIn, X, and Instagram…