---
title: Design Intelligence Kit
category: product
entity_type: skill-kit
price: $38
canonical: https://forgehouse.ai/ai-kits/design-intelligence-kit/
lang: en
hreflang_alt: https://forgehouse.ai/tr/ai-kitler/design-intelligence-kit/
last_updated: 2026-06-20
---

# Design Intelligence Kit

> Decision-led design craft, not AI slop.

Three skills that turn design from guesswork into decisions: a sector knowledge base, a neuroscience-driven craft loop, and a Claude Design workflow.

**Price:** $38, one-time purchase, no subscription.

## What's inside
- design-intelligence
- design-craft
- claude-design

## How to use
Pull sector data, run the craft loop to iterate, and use the Claude Design workflow to produce and polish.

## From taste to decision, pass by pass
The kit is not an AI button you press. It is the craft loop we run when a design has to look decided, not generated: sector data sets the direction, craft verifies it, contrast is measured.
1. Design Intelligence starts with sector data: palette, font and UI style chosen from what works in the field, not from taste.
2. Claude Design Integration turns that decision into a real layout from plain language.
3. Design Craft runs three passes: structural composition first, then visual verification against the rendered screen, then a self-audit.
4. Contrast is measured against WCAG, not eyeballed, so 'looks fine' becomes a number that passes.
5. The loop repeats until the section reads as decision-led, not AI slop.

Craft run
- INTELLIGENCE: palette + font set · from sector data
- DRAFT: generated · plain-language to layout
- CRAFT: 3 passes · compose / verify / audit
- CONTRAST: measured · WCAG pass

## Field case, A section that looked amateur with no obvious reason (2026)
Problem: A page section felt cheap and off, but no one could name what was wrong. The instinct was to add more: another gradient, a bigger shadow, a new font.
Fix: Design Intelligence reset the palette and font from sector data instead of taste. Design Craft then ran the three-pass loop: it composed the structure, verified it against the actual rendered screen, and self-audited. Contrast was measured against WCAG rather than guessed.
Result: The section moved from a vague 'something is off' to a decision you could explain line by line. Every color and type choice traced back to a reason, and the contrast passed as a measured value, not a hope.

## Who it is for
- Agencies and designers who keep hearing “this looks cheap” and need a repeatable fix
- Builders who want every palette, font and layout decision backed by sector data
- Teams who design in natural language but need to ship real, production-grade code

## Proof
How we keep client UI premium and on-brand: research-grounded, contrast-verified, not generated guesswork.

---
Maker: Can Davarcı, https://candavarci.com.tr