AI Social Media Content Factory with Multi Platform Publishing
N8N
Project's Category
Intro
This one pulls prompt and schema instructions from an external source, has an AI agent write and format a post for each platform, generates a matching image, and waits for human approval before publishing anywhere.
Problems
Posting the same idea across Twitter, Instagram, Facebook, LinkedIn, Threads and YouTube Shorts meant reformatting the same content six different ways by hand, and there was no single point where someone could review a post before it went out everywhere at once.
Solutions
We built an n8n system where a router agent pulls a system prompt and platform schema from an external source, hands it to a content creation agent with web search and chat memory, and generates a post plus a matching image for each target platform. Before anything gets published, the draft is packaged into an HTML approval email and held until a human approves it. Once approved, a publishing router sends the right format to X, Instagram, Facebook, LinkedIn, Threads and YouTube Shorts, saves a copy to Google Drive, and emails a summary of what went out.
Final Thoughts
Keeping the approval step in the middle instead of at the very end means nothing goes out that a person hasn't actually seen first. Pulling the prompt and schema from an external source instead of hardcoding it means the whole thing can be repointed at a different brand voice without touching the workflow itself.
AI Social Media Content Factory with Multi Platform Publishing
N8N
Project's Category
Intro
This one pulls prompt and schema instructions from an external source, has an AI agent write and format a post for each platform, generates a matching image, and waits for human approval before publishing anywhere.
Problems
Posting the same idea across Twitter, Instagram, Facebook, LinkedIn, Threads and YouTube Shorts meant reformatting the same content six different ways by hand, and there was no single point where someone could review a post before it went out everywhere at once.
Solutions
We built an n8n system where a router agent pulls a system prompt and platform schema from an external source, hands it to a content creation agent with web search and chat memory, and generates a post plus a matching image for each target platform. Before anything gets published, the draft is packaged into an HTML approval email and held until a human approves it. Once approved, a publishing router sends the right format to X, Instagram, Facebook, LinkedIn, Threads and YouTube Shorts, saves a copy to Google Drive, and emails a summary of what went out.
Final Thoughts
Keeping the approval step in the middle instead of at the very end means nothing goes out that a person hasn't actually seen first. Pulling the prompt and schema from an external source instead of hardcoding it means the whole thing can be repointed at a different brand voice without touching the workflow itself.
Building the automation systems that handles business processes.
Clear Communication
Reliable
Professional
Reliable Quality
Building the automation systems that handles business processes.
Clear Communication
Reliable
Professional
Reliable Quality
