Design thinking
Structured thinking skills for AI agents: brainstorming, decision support, systems thinking, and strategy. Built for Claude Code & Cursor.
npx -y skills add ysskrishna/ai-agent-skills --skill design-thinkingAssembled from the repository path, not quoted from the project. Check it against their README if it does not work.
One thing to look at
- 3 stars3 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.
What its author says it does
Copied from the file, not written here
Use this skill when the user asks for design thinking (including naming it or directing use/apply/run with obvious misspellings; decisive) or wants human-centered exploration—empathizing with needs, framing the problem, ideating, prototyping intent, and defining what to learn next. Use for HCD, service or UX concept sprints, how-might-we style discovery before build, or reframing from user evidence, even with messy context. Skip when the spec is fully frozen and they want no discovery, or when the task is code-only maintenance with no user problem framing requested.
The file declares its own license as MIT. That is the author’s claim about this one file, and it is not the same thing as the license GitHub reports for the repository, which is listed with the other numbers below.
SKILL.md
3.3 KB, as published. Nobody here has run it
Design Thinking
Fall in love with the problem, not the first solution. End with what to learn next, not just ideas.
How to run it with this skill: one clearly headed section per stage in this order: Empathize → Define → Ideate → Prototype → Test plan.
Setup (run before starting)
In one short block:
- Design challenge — who is affected and in what situation?
- Default pass — Empathize → Define → Ideate → Prototype → Test plan (state this line)
If users, constraints, or success signals are missing, ask at most 3 questions in one message, then proceed. Note any remaining gaps or working guesses in plain language (no bracket tags in Setup).
If Empathize is thin (no real user input), say so honestly in Define and keep the POV narrow instead of inventing research.
The Stages
Empathize
Who — primary user or stakeholder (facts from user vs [INFERRED]). Jobs / pains / gains — what they are trying to do and what hurts. Context — when/where the need shows up.
No fabricated quotes; paraphrase only what the user supplied.
Define
Insight statement — non-obvious tension connecting pains and context. Point of View (POV) — "[User] needs [verb] because [insight]." How Might We (HMW) — 2–3 well-scoped questions opened by the POV.
Ideate
Quantity + variety. Use HMW as prompts. Tag ideas desirable / feasible / viable as hypotheses (not proven). Produce a substantive list (no fixed count unless the user specifies one).
Prototype
Describe low-fidelity artifacts: paper flow, roleplay script, landing smoke test, clickable sketch. For each: Purpose: what question does this answer? Fidelity note: one line placing the artifact on a sketch-only vs interactive spectrum (no low/medium labels).
Test plan
Learning goals — what would convince you the idea is wrong? Participants / sample (or [TBD]). Signals — behaviors or metrics to observe. Next iteration — what changes if results are mixed.
Execution Rules
- Define must reflect whatever Empathize actually contains; do not invent field research.
- Do not collapse Ideate into a single solution before Prototype.
- Test plan must include falsifiable signals.
Checklist (verify before responding)
- Setup: design challenge + default pass
- Empathize distinguishes fact vs
[INFERRED] - POV + HMW before Ideate
- Ideation is substantive; desirable / feasible / viable tags used
- Prototype states purpose and a one-line fidelity note (no low/medium labels)
- Test plan has learning goals and signals