← Claude Code
Riley Brown
Fable 5 Replaced My Designer (Here's How)
⏱ 24 min video · 3 min read6 Jul 2026Worth watching
TL;DR
Riley Brown demos an AI-native design workflow using Claude Code connected to Paper (paper.design) via MCP, automating the creation of YouTube thumbnails, Instagram graphics, slide decks, and deployed landing pages entirely through prompts. The video is a practical walkthrough showing how to set up the integration and use it across multiple real design use cases without a human designer.
Key points
1
Paper (paper.design) is a Figma-like canvas app built natively for AI agents, controllable via an MCP server that Claude Code connects to through a simple setup prompt.
2
Claude Code can scrape YouTube thumbnails from any channel and import them directly into a Paper board using a custom skill, enabling rapid thumbnail ideation by swapping faces and text.
3
Paper has built-in AI image generation that supports multi-image reference prompts, letting users replace characters, icons, and text in thumbnails iteratively.
4
Instagram graphics with gradient overlays and editable text layers can be generated and iterated entirely through Claude Code prompts, with assets referenced by copying Paper asset links.
5
A full 6-slide sponsor pitch deck was built by Claude Code using Perplexity branding scraped via Firecrawl, then deployed as a live website on Vercel in a single follow-up prompt.
Actionable insights
→
To connect Claude Code to Paper, go to paper.design, click 'Using Agents', copy the MCP link at paper.design/docs/mcp, then paste it into a Claude Code chat and ask it to set up the MCP server connection.
→
Use frames in Paper to group multiple images and share a single frame link with Claude Code instead of copying individual asset links, saving time when referencing multiple photos.
→
Use the Firecrawl API as a scraper skill inside Claude Code to pull styling from any website (e.g. Perplexity) and replicate that brand style in presentations or landing pages built inside Paper.
→
When iterating on AI-generated graphics, select the target image first before shift-clicking reference images so Paper knows the correct output aspect ratio.
→
Set Claude Code effort level to 'low' for quick connection tests and simple tasks, and switch to 'high' only for complex outputs like full decks to conserve Fable tokens.
Notable quotes
“I don't open paper anymore. Claude does. That's kind of fire.”
“This video is not sponsored by Claude, nor is it sponsored by paper. I just really genuinely use these tools a lot especially for thumbnails and Instagram graphics.”
Worth watching?
✅
Worth watching the full video?
Watch if you want to see the live screen recording of this workflow in action — the key setup steps and use cases are all captured here, but the visual demonstration of Claude autonomously building graphics in real time adds useful context you cannot get from reading alone.
Saved you some time? The creator still deserves a like.
Found this by searching? That's the hard way.
summree watches the AI channels you follow and sends you the summary the moment a video drops. Nothing gets past you. From £4/month.
Try it free