I want a fully-automated n8n workflow that plans, creates, and publishes social-media content for Facebook, Instagram, and YouTube. The flow should do everything from idea generation to live posting with zero manual touch points. Here’s how I picture it working: • A daily trigger (cron or webhook) kicks things off. • ChatGPT receives a brief prompt template, researches a trending topic or keyword I specify, and returns post copy, hashtags, and an optional video script. • If the post needs a visual, the workflow calls a generative image tool (DALL·E, Midjourney, or another you recommend) and pulls the file back into n8n. • n8n then formats the result for each platform—caption length for Instagram, description and tags for YouTube, etc.—and publishes through the official APIs. • Logs, error handling, and a simple Google Sheet or Airtable archive keep track of what went out and when, so I can audit results later. Deliverables I expect: 1. An exported n8n workflow (.json) that I can import into my own instance. 2. A short README explaining required environment variables (API keys, tokens) and how to customise prompts, schedules, and platforms. 3. A quick Loom or written walkthrough that proves a successful end-to-end run for text, image, and video posts. If you can add optional branches for Twitter or LinkedIn I’m open, but the must-haves are Facebook, Instagram, and YouTube. ChatGPT is my preferred model; feel free to propose Gemini for comparison, but the final build should default to GPT. I’m looking for clean, maintainable node organisation and thoughtful fallbacks (e.g., if an image generation call fails, still push a text-only post). Let me know how you’d structure this and any similar automations you’ve built.