# Design Sprint

> Run a structured 5-day design sprint from problem framing to prototype testing. Use at the start of a project for ideation, solution selection, rapid prototyping, and concept validation.

- Skill: `talgacapri/design-sprint` (Agent Skill)
- Install (CLI): `npx skillmds@latest add talgacapri/design-sprint`
- Raw SKILL.md: https://api.skillmd.com/api/skills/talgacapri/design-sprint/raw
- Safety review: pending
- Works with: Claude Code, Claude.ai, OpenAI Codex
- Category: Coding & Dev Tools
- Author: talgacapri (https://skillmd.com/u/talgacapri)
- Updated: 2026-09-17
- Page: https://skillmd.com/skills/talgacapri/design-sprint

---


# Design Sprint

Use this skill at project kickoff to reduce risk before full build.

## What this skill does

This skill runs a compact 5-day sprint process:
- Understand and frame the problem
- Generate and compare solution options
- Decide what to prototype
- Build a realistic prototype
- Test with target users and synthesize learnings

## Best triggers

Activate for requests like:
- "run a design sprint"
- "ideation workshop"
- "validate this idea"
- "test concept before build"

## 5-day structure

### Day 1: Map
- Define long-term goal
- Map user journey and critical moments
- Pick the target area

### Day 2: Sketch
- Create multiple solution sketches
- Keep ideas concrete and comparable

### Day 3: Decide
- Review options with clear criteria
- Select one concept (or a testable combination)
- Write storyboard for prototype

### Day 4: Prototype
- Build a realistic enough prototype for learning
- Focus on key path, not full feature depth

### Day 5: Test
- Run user tests
- Capture reactions, confusion points, and value signals
- Convert findings into go/no-go and next steps

## Output format

Return:
1. Sprint plan and participant roles
2. Daily agenda with deliverables
3. Decision criteria and scorecard
4. Prototype scope definition
5. Test script and synthesis template
6. Recommendation: ship, iterate, or stop

## Guardrails

- Time-box decisions to preserve momentum.
- Prototype only what is needed to learn.
- Prioritize learning velocity over polish.
- End with concrete decisions, not just observations.