Designing With AI: Claude, Codex, Figma | Full Guide

You are helping me turn one specific YouTube video into real, durable learning.

Source video:
- Title: Designing With AI: Claude, Codex, Figma | Full Guide
- URL: https://www.youtube.com/watch?v=j_ZPV10bu54
- Topic: Interfaces + Open Design
- My current learning frame: Generate one screen in Claude using your design-system skills plus two or three Mobin reference screenshots, push it through Figma, then bring it into Codex to make bulk edits and compare the token cost of each step.
- Why this matters: This is the strongest new bridge between the learning atlas, design taste, and real product-building workflow.

Transcript anchors from this exact video:
- 0:00 / Evidence 1: "Today we're breaking down the full AI design workflow from start to finish. We'll look at the current state of AI tools in Figma, how to set everything up, when to use Claude, Codeex, Claude Design, Figma, and..."
- 7:56 / Evidence 2: "of it inside of cloud code. So, it's not like a onetoone both produce the exact same code. However, Codex uses about three to four times fewer tokens for the same work as Claude. What this means is..."
- 16:34 / Evidence 3: "reasons. One, because Google Stitch, we can't train it on our on our design system the way that we would expect. We can't paste in a Figma file here and build skills around our design system. That's a..."
- 32:50 / Evidence 4: "inside claude code and codec. Let's run the exact same prompt we've been working with. It's not just about comparing outputs but your AI workflow might change as part of it once we look at tokens and how..."
- 38:14 / Evidence 5: "we're talking about how many tokens it took to do everything so far is because Claude had to build a design from scratch, but Codex just brought in a design from Figma. So, it's a little bit of..."
- 40:18 / Evidence 6: "we're at a point where we understand some of the key tools in the AI design space right now. What that workflow could look like depending where you are in your design journey, but I want to talk..."
- 77:42 / Evidence 7: "is going to help us inform claude code. So using the screenshot uh attached or the uh reference example attached along with the uh variables type styles and component skills skills. Please build please uh build a page..."

Video-aware target:
- Prompt lane: Design system
- Mechanism to extract: Extract how the video turns visual references or component systems into usable constraints for agents.
- Artifact to produce: A design-system adoption brief with source references, tokens/components, agent handoff rules, and visual QA checks.
- Artifact must include: references; tokens/components; handoff artifact; implementation rule; visual QA

Your task:
1. Use the transcript anchors above as the primary source packet. If you add outside context, label it clearly as outside context and keep it secondary.
2. Create a source-check table with columns: timestamp, claim, transcript support, what the demo proves, confidence, and what still needs verification.
3. Extract the actual teachable mechanism from the video: Extract how the video turns visual references or component systems into usable constraints for agents. Do not invent claims that are not supported by the title, lesson frame, or transcript anchors.
4. Build a reusable learning artifact: A design-system adoption brief with source references, tokens/components, agent handoff rules, and visual QA checks.
5. Include:
   - a plain-English definition of the core idea
   - a diagram or structured model using this sequence: Reference -> Tokens -> Components -> Usage rules -> Agent prompt context -> Implementation -> Visual QA
   - answers to these source questions: What design source is reused? | How is it translated into agent context? | What review catches generic output?
   - 3 concrete examples that apply the video idea to real agentic work, such as Figma-to-shadcn workflow; design.md brief; UI reference library remix
   - 2 failure modes the video helps prevent, chosen from the transcript evidence and these likely risks: copying visuals without rules; generic generated UI; no visual QA screenshot pass
   - a checklist for the next real workflow, focused on: references, tokens, components, handoff, QA
   - one practical exercise with a clear done signal: Turn one screen reference into five constraints a coding agent must follow.
6. Add a "learning transfer" section: what changes in my workflow tomorrow if I actually learned this?
7. Add a "source check" section that cites which transcript anchor supports each major takeaway.

Quality bar:
- Make this specific to "Designing With AI: Claude, Codex, Figma | Full Guide", not a generic Interfaces + Open Design essay.
- Cite transcript anchors for every claim about design context, UI generation, preview, critique, or handoff.
- Prefer operational examples, failure modes, and reusable artifacts over broad definitions.
- Call out uncertainty instead of smoothing over weak evidence.
- Avoid these generic drifts: generic design inspiration; component lists without usage rules; no screenshot review.
- If evidence is weak or missing, stop and say what transcript segment or timestamp needs review instead of guessing.
- Finish with a concise artifact I could paste into my learning app.