/crisp-audit
A full, scored review of your design.
(What it does)
The full CRISP evaluation. /crisp-audit reviews a design the way a senior product designer would. It starts with a thirty-second first impression, then scores each of the five dimensions: Contextual, Responsive, Intelligent, Seamless, and Powerful. Every problem it finds is named, rated from P0 (blocks the user) to P3 (minor polish), and paired with a specific fix.
You get a scorecard, a prioritised fix list, a comparison with Stripe, Linear, Notion, Asana, and Slack (or the benchmarks in your .crisp.md), and recommendations for reaching world-class quality.
(When to use it)
Use it at decision points, not while you are still iterating:
- Before an important screen ships, as the last structured check
- Milestone reviews, where you need a grade and a scorecard you can justify
- You have inherited a UI you did not build and need an honest assessment
/crisp-reviewkeeps flagging the same screen and you need the full picture
(What you get)
- Grade A–F with reasons
- A score for each of the five CRISP dimensions
- Problems rated P0–P3 by severity
- Prioritised fix list
(Example output)
Illustrative example only. This is not a real audit.
(How the audit runs)
- 30-second scan: first impressions, strengths, red flags, and a provisional grade
- Dimension scoring: each of the five dimensions checked against clear signs of failure
- Severity rating: every problem rated from P0 (blocks the user) to P3 (polish)
- Benchmark comparison: how the design compares with world-class products
- Action plan: fixes ordered by impact on users, plus quick wins
(Common questions)
How is this different from /crisp-review?
Depth and purpose. /crisp-review is a thirty-second check: one grade and the top three issues. /crisp-audit scores every dimension separately, rates every problem by severity, compares the design with named products, and ends with a plan. Use review while you iterate and audit at milestones.
What do the P0–P3 ratings mean?
P0 blocks the user completely, such as an empty state with no way forward. P1 is major friction the user can work around but should not have to. P2 is a noticeable drop in quality. P3 is minor polish. The severity scale turns an audit from a list of opinions into an ordered plan.
What does it benchmark against?
Stripe, Linear, Notion, Asana, and Slack by default. Each is chosen for a specific strength, from Stripe's progressive disclosure (showing advanced options only when they are needed) to Linear's keyboard-first speed. If .crisp.md lists your own benchmarks, those are used instead.
Can it audit a screenshot?
Yes. A screenshot, a Figma link, or a written description all work. The more detail you give, the more precise the findings. With code or a live UI, the audit can point at exact elements rather than general areas.
(Install)
Installs all fourteen CRISP skills and detects which AI coding tool you use. Run /crisp-teach once per project, before the other skills.
(Related skills)