Developer Tools2025

DesignSkills

Design thinking skills for AI-assisted product development. A structured toolkit that guides teams from problem framing through usability evaluation, built for Claude Code.

8 skills shippedOpen-sourceApache 2.0
Claude CodeDesign ThinkingCLIOpen SourceView on GitHub
DesignSkills

TL;DR

Built an open-source toolkit of 8 design thinking skills for Claude Code that guides AI-assisted teams from problem framing to usability evaluation — so developers ship user-centered products, not just code.

Get started with one command

npx add-skill Abhsin/DesignSkills

Overview

Role

Creator & Sole Developer

Platform

Claude Code CLI

License

Apache 2.0

The Problem

AI builds fast, but skips the thinking

AI coding agents can generate entire applications in minutes, but they jump straight to implementation. There's no structured way to ensure the right problem is being solved, the right users are being designed for, or the right assumptions are being validated. Teams end up with technically sound products that miss the mark on user needs.

Key Insights

01

Design thinking methodology can be encoded as structured, repeatable skills that AI agents follow

02

Artifacts from each phase should persist as markdown files that both humans and AI can reference

03

A full pipeline approach (problem → persona → assumptions → scope → docs → build → evaluate) prevents premature implementation

The Solution

8 skills across 4 phases

Each skill is an interactive, guided process that extracts structured thinking from the user and generates design artifacts in a consistent format.

/problem-framing

Converts vague ideas into clear, actionable problem statements

Pre-Planning
/user-modeling

Develops behavior-based personas from research data

Pre-Planning
/assumption-mapping

Identifies and prioritizes risky assumptions early

Pre-Planning
/solution-scoping

Defines MVP boundaries and feature prioritization

Pre-Planning
/prd-generation

Creates lean PRDs with user stories and acceptance criteria

Documentation
/ux-specification

Defines interaction flows, screens, and component behavior

Documentation
/prompt-export

Converts specifications into sequenced build prompts

Prompts
/heuristic-evaluation

Conducts usability reviews using Nielsen’s heuristics

Evaluation

Workflow

Two paths to shipped product

Full Pipeline

Problem framing → User modeling → Assumption mapping → Solution scoping → PRD → UX spec → Prompt export → Build → Heuristic evaluation

Fast Iteration

Problem framing → Solution scoping → PRD → Prompt export → Build → Heuristic evaluation. Skips intermediate steps for rapid MVP validation.

Architecture

Structured artifact output

Each skill generates numbered markdown files in a design/ directory, creating a persistent record of design decisions that both team members and AI agents can reference throughout development.

project/

design/

01-problem-statement.md

02-user-personas.md

03-assumption-map.md

04-solution-scope.md

05-prd.md

06-ux-spec.md

07-heuristic-review.md

prompts.md

src/

What's Next

Expanding the toolkit

Ideation Skills

Problem reframing and creative brainstorming tools

Design Critique

Structured critique frameworks for design review

Usability Scripts

Automated usability testing script generation

Reflections

What I learned

Building DesignSkills taught me that design methodology doesn't have to be a heavyweight process. When encoded as interactive skills, design thinking becomes a natural part of the development workflow rather than a separate phase that teams skip under time pressure. The key was making each skill self-contained yet composable.