Skip to content
Agentmart
Submit a tool

design-critique

A heuristic critique of a screen or flow against the Nielsen ten usability heuristics plus the studio rubric, from a screenshot, a file or a URL.

Skill by adityareddy-dev

IndexedMIT★ 41
Write a review

Install

Works with Claude Code, Claude apps, Codex, Cursor, Gemini CLI, GitHub Copilot, VS Code, OpenCode, Goose, Windsurf. Read the repository before you give any tool access to your data.

npx agentmarket install skill/adityareddy-dev-design-critique

Asks which of your tools to add it to (Claude Code, Codex, Cursor, VS Code, Windsurf, Gemini CLI, Claude Desktop) and shows the change first. See https://agentmart.dev/cli.

Get the weekly diff

The weekly diff: one email a week with new agent tools, notable updates and grade changes, plus security alerts for tools you follow. Double opt-in, one-click unsubscribe.

We use your email only for this newsletter. Privacy

About

A heuristic critique of a screen or flow against the Nielsen ten usability heuristics plus the studio rubric, from a screenshot, a file or a URL. Use when you want a usability read, not a redesign.

Use cases: design

No reviews yet. Installed it? Reviews come from verified installers only. How reviews work

Permissions

What this skill can do and what it asks for, from its SKILL.md. Checked Oct 25, 2026.

Tools
No tools found
Secrets
None asked for
Install scripts
None found

Network hosts it talks to 1

  • codeload.github.com · found in install command

Files it touches

  • ~/.claude/skills/design-critique

What's inside

Skill name
design-critique
Description
A heuristic critique of a screen or flow against the Nielsen ten usability heuristics plus the studio rubric, from a screenshot, a file or a URL. Use when you want a usability read, not a redesign.
License
MIT
Instructions
30 lines in SKILL.md · view on GitHub

Files (1)

  • SKILL.md

SKILL.md (first lines)

Critique a design as the Design Manager. No roughs, no pick, no polish: the studio looks, names what works and what fails, and says what to change.

**Deliverable and standard.** A critique report. Scored on two lists: the Nielsen ten (visibility of status, match with the real world, user control, consistency, error prevention, recognition over recall, flexibility, minimal design, error recovery, help) each rated pass, weak or fail with the evidence, and the studio rubric's eight axes. The standard is whichever matches the input, else the house style.

**Desks.**

| Desk | Agent | Tier | Does |
| --- | --- | --- | --- |
| UX Designer | `desk:ux-designer` | senior | the ten heuristics, with the element and the fix for each weak or fail |
| Content Designer | `desk:content-designer` | builder | labels, errors, empty states and microcopy against the copy standard |
| Creative Director | `desk:creative-director` | director | the rubric scores and the three changes that matter most |

**Inputs.** A screenshot (the default mode), a set of screenshots for a flow with `--flow`, an HTML file, or a URL. Screenshot mode is the common case: nothing is rendered, the desks read the images. A file or URL is rendered by proof at the sizes first so the critique is of the real thing, and the proof failures are listed under the heuristics as evidence.

**Proof sizes.** Screenshot mode: none. File or URL: `--sizes`, else `phone,desktop`.

**Output** under `.desk/out/<id>/`:

- `critique.md`: a one-paragraph read, the heuristics table (heuristic, rating, where, fix), the rubric scores, the copy findings, then "Three things to change first".
- `renders/` when a file or URL was rendered.

**Run.** Read `${CLAUDE_PLUGIN_ROOT}/skills/desk/SKILL.md` now and follow it from step 0 to step 7 as the Design Manager with the desks above. Command name for state and events: `design-critique`. Changes to the loop for this command:

- Skip roughs, pick, polish and sign-off as gates. Phase order: intake, render (file or URL only), critique (UX Designer and Content Designer in parallel), score (Creative Director), report.
- The Creative Director's scores are a reading, not a gate: under-bar does not loop. The report carries them as they are.
- In screenshot mode the brief's Sizes line says "screenshot" and state records `proof: []`.

The ask: **$ARGUMENTS**

From adityareddy-dev/deskstudio, MIT.

More from adityareddy-dev/deskstudio

Signals

Stars
41
Forks
41
Open issues
10
Last commit (this folder)
Oct 20, 2026 (9 days ago)
Last release
Oct 23, 2026 (5 days ago)

From GitHub (fetched Oct 24, 2026).

Details

License
MIT
First seen
Oct 24, 2026
Last checked
Oct 24, 2026

More from adityareddy-dev

Cite this page

design-critique skill. Agent Mart. https://agentmart.dev/skills/adityareddy-dev-design-critique (last updated 2026-10-24)