agentleFS
gnegreiros-uxgnegreiros-ux/agentica-design-system/.agents/skills/ux-pattern-review/SKILL.md

ux-pattern-review

Skill2 starsChanged 46 days agoApache-2.0View on GitHub

Before publishing a new or UX-relevant-modified Agentica component, look up the reference UX patterns from the 5 source-of-truth references, present them to the human with direct links for approval, then document the decision across 6 surfaces (guideline, code, story, site, ADR, GitHub Projects). Use when creating a new component or making a UX-relevant change (new variant/state, validation logic, error/help display, interaction) to an existing one.

1 other project uses a version of this.

Adopted by

Most recently taken up by

The file

We split this skill into the rules it gives an agent

5 rules. About 0 are common, 3 are distinctive, and 2 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈3

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 2

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Very specific:

    Before publishing a component — new or modified in a UX-relevant way — present to the human the UX patterns suggested by the 5 reference …

    line 16
  2. Very specific:

    Read the type → priority sources matrix in governance/rules/ux-patterns-sources.md.

    line 42
  3. Distinctive:

    Look up the priority source(s) for this component type (fetch the page if you have web access.

    line 43
  4. Distinctive:

    Always include NN/g as the usability baseline.

    line 45
  5. Distinctive:

    If in doubt about the impact (e.g.

    line 57

≈ “Common” is an estimate: at least 1 in 5 of the 372,429 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Skill: ux-pattern-review

Cross-tool port of .claude/skills/ux-pattern-review/SKILL.md (Claude Code's native implementation) — see governance/ai-skills-reference.md for the full picture of what Agentica's skills automate. This file follows the open Agent Skills format (agentskills.io) so Codex CLI and GitHub Copilot CLI load it automatically from .agents/skills/; content is identical in substance to the Claude Code version.

Objective

Before publishing a component — new or modified in a UX-relevant way — present to the human the UX patterns suggested by the 5 reference sources (see governance/rules/ux-patterns-sources.md), with direct links, so they can judge and approve which ones to apply. The decision is then documented everywhere.

The human always has the final word. This skill proposes, the human decides.

When to run this skill

  • On the creation of a new component (always).
  • On a relevant modification of an existing component: new variant/state, change to validation logic, error/help display, an interaction, or adding a type.
  • Not necessary for: contrast fix, typo, variable rename, refactor with no behavior change.

Process

Step 1 — Frame it

Identify:

  • The component involved and its type (field, action, feedback, container, icon, navigation, data…).
  • The nature of the change (creation vs. modification; which UX aspect is affected).

Step 2 — Consult the sources (hybrid)

  • Read the type → priority sources matrix in governance/rules/ux-patterns-sources.md.
  • Look up the priority source(s) for this component type (fetch the page if you have web access; otherwise use what you already know of that source and say so).
  • Always include NN/g as the usability baseline.

Step 3 — Present the suggested patterns

Produce a table (format below) listing each candidate pattern with: name, source + direct link, problem it solves, default recommendation. Explicitly cover the review checklist questions: states, error display, help text, validation timing, required markers, progressive disclosure, dark patterns.

Step 4 — Wait for human approval

  • The human checks off the patterns to apply (✅) and rejects the others (❌), with justification.
  • Do not build or publish anything before this approval.
  • If in doubt about the impact (e.g. critical action, sensitive data): escalate.

Step 5 — Document on the 6 surfaces

Propagate the Pattern Decision Record (see governance/rules/ux-patterns-sources.md):

  1. Guideline guidelines/components/<comp>.md → section ## UX Patterns Reference.
  2. Code components/agtc-<comp>.js → "WHY" header comment block + links.
  3. Storybook components/agtc-<comp>.stories.js → parameters.docs.description.component.
  4. Site → node site/build.js.
  5. ADR for the component's implementation → list of applied patterns.
  6. GitHub Projects → reflect the work item (status, domain) — see ADR-069.

Output format — suggested patterns

## UX pattern review — <component>

Type: <type> · Nature: <creation | modification: aspect affected>
Sources consulted: <list with links>

| Pattern | Source (link) | Problem solved | Recommendation | Decision |
|---------|---------------|-----------------|----------------|----------|
| Inline error under the field | [NN/g — error handling](https://www.nngroup.com/articles/design-pattern-guidelines/) | Where to show the error | Recommended | ☐ ✅ / ☐ ❌ |
| onBlur validation (not onChange) | [IxDF — forms](https://ixdf.org/literature/topics/ui-design-patterns) | When to validate | Recommended | ☐ ✅ / ☐ ❌ |
| Required marker `*` + aria-required | [IxDF — required fields](https://ixdf.org/literature/topics/ui-design-patterns) | Signal required fields | Recommended | ☐ ✅ / ☐ ❌ |
| … | … | … | … | ☐ ✅ / ☐ ❌ |

### Open questions for the human
- [validation timing, edge cases, trade-offs]

### Awaiting approval
> Check off the patterns to apply. No publishing before your decision.

What this skill does NOT do

❌ Invent a pattern not from the reference sources
❌ Decide alone which patterns to apply
❌ Build or publish a component without human approval
❌ Skip propagation to the 6 surfaces
❌ Bypass quality-gate's UX pattern review requirement

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.