Products/Developer Tools / AI Design Tools/Taste Lab

Taste Lab

Design DNA for Agents

Developer Tools / AI Design Toolsdeveloperai-design

Our Take

Taste Lab is doing something I haven't seen anyone else attempt — it's teaching AI agents design taste, not just stealing design tokens. Linear, Stripe, Anthropic, and Pipeline are already using it to give their agents context on why a design decision matters, not just what the hex code is. The tool maps 20 measurement categories to a "Taste DNA" framework that explains the trade-off behind every choice — the difference between an agent that copies pixels and one that actually understands them. If you're building with Cursor, Windsurf, or any agent tool, this is the missing piece.

turns any URL into a complete design context for your AI agent. Extracts Design Map, Taste DNA, and the reasoning behind every design choice.

Key Features

Extract Measurements - Senior Design AI extracts precise objective design measurements from a webpage, Detect Patterns - Principal Design Engineer detects systematic rules in measurements, Infer Taste - Ultimate Design Critic derives deliberate trade-offs the designer made, Observer - Lead Critic filters taste analysis with quality gate, Design Map - 20 measurement categories with exact px/hex/ratio values, Taste DNA - 4 principles each with Trigger, Decision, Reason, Evidence, and Trade-off, Output as .md and .json files formatted for each AI tool, Works with Cursor rules, Windsurf rules, Claude Code CLAUDE.md, GitHub Copilot instructions, Bolt prompt, Antigravity GEMINI.md, v0 CSS, Figma Make CSS

Problem It Solves

When an agent has only tokens, it copies the numbers without understanding them. An agent with taste knows what each decision meant and can make the right call on a page it's never seen.

Target Customer

AI agent developers building with Cursor, Windsurf, Claude Code, GitHub Copilot, Bolt, v0 by Vercel, Figma Make, and Lovable

Use Cases

Reverse engineer any website's design taste, Extract complete design token sets, Understand design reasoning behind choices, Give AI agents design context for better decisions

Differentiator

Focuses on decisions × trade-offs rather than just extracting tokens. Explains the why behind each design choice.

Why Now

AI agents need to understand design context, not just copy numbers. When an agent encounters a new page, it needs to know what each design decision meant.

Traction

Customers Mentioned: Linear, Stripe, Anthropic, Pipeline

Key Facts

The people behind Taste Lab

Sen Lin

profile

Vibe builder

LinkedIn GitHub

Links

Website GitHub Source: product-hunt

Want products like this in your inbox every morning?