Repurpose article into X thread and LinkedIn post

How it works

1. 1.Form trigger exposes a required textarea for the article/transcript and an optional voice notes field
2. 2.The agent prompt injects the source (whitespace-normalized, sliced to 9,000 chars) with rules banning invented facts and outputting [NEEDS-HUMAN] for thin or unreadable input
3. 3.The model generates a numbered X thread (max 270 chars per tweet, hook first, CTA last) and a skimmable LinkedIn post
4. 4.The form completion node shows the agent output to the user under the title 'Your posts are ready'
5. 5.The source is explicitly treated as content, not instructions, mitigating prompt-injection via pasted text

Use cases

• Turning blog articles or newsletters into ready-to-publish X threads
• Converting video or podcast transcripts into LinkedIn posts
• Helping creators redistribute one long asset across two platforms without rewriting

Requirements

• An OpenAI credential configured in n8n for the gpt-4o-mini chat model node
• The form must be activated or tested via its production/test URL before use
• Input source should contain roughly 200+ useful words or the agent will decline to generate

Setup and testing in n8n

1. Open the form URL and confirm both fields render with the required flag enforced
2. Submit a substantial article and verify the completion page shows both delimited output sections
3. Submit a very short or garbage text and confirm the output starts with [NEEDS-HUMAN] plus a reason
4. Cross-check that all facts and numbers in the posts appear in the submitted source

Service Value

Ideal as a smart automation service combining integrations and AI to produce ready-to-use results.