Skip to content

Improve UI skill

Audit one product surface from proven design evidence and write selected implementation plans.

Community · ibelick
Authoribelick
LicenceMIT
Sourceibelick/ui-skills · SKILL.md on GitHub
Project websitewww.ui-skills.com
Works inClaude Code · Codex, Gemini CLI, Cursor, Copilot
Read from the source on

improve-ui audits an existing interface against the design decisions that actually govern the selected surface.

It is read-only on product source. Findings need contract, runtime and correction evidence; the report stops at three, and plans are written only for user-selected changes.

From the skill's SKILL.md: # Improve UI

This is a community skill by ibelick, not checked by us or by Anthropic. Skills can tell the agent to run code, read files or call the network. Before installing, read its SKILL.md and every script in the folder, and look for anything that does not match what it says it does.

What it does

  1. Trace one coherent surface through routes, components and resolved styles.
  2. Prove and try to falsify each candidate before reporting it.
  3. Write self-contained plans under design-plans for selected findings.

What you get

  • Design-language evidence and up to three supported findings.
  • Selected implementation plans for another executor.

When to use it

  • Review design drift without replacing an established product identity.
  • Prepare a source-grounded design handoff.

When not to use it

  • Expecting this audit to implement changes or automatically include functional and accessibility findings.

What it needs

  • Read access to the relevant source and governing design documentation.
  • Rendered evidence only when supplied or explicitly requested; no required MCP service or package.

Files in the skill

File or folderWhat it is for
SKILL.mdEvidence gates, scope boundaries and report structure.

Licence as found in the repository on 2026-11-17: MIT.

Example prompts

Type one of these after installing; naming the skill makes sure it is used.

  • Audit this checkout surface against its current tokens; do not modify product source.
  • Try to falsify these UI findings before recommending a change.
  • Turn my selected finding into a self-contained plan under design-plans.

How to install improve-ui

Read-only audit

Copy the single skill folder. The optional UI Skills registry MCP is a discovery service, not a runtime requirement.

The skill excludes accessibility and semantic findings unless requested. Use a dedicated accessibility review when that is the task.

A skill is a folder with a SKILL.md file. You install it once; the agent reads its short description at the start of each session and loads the full instructions only when a request matches. Pick the place you use:

Claude Code: copy the folder into your skills folder

Put the skill's folder in ~/.claude/skills/ to use it in every project on this computer, or in .claude/skills/ inside a repository to share it with everyone who works there (commit it).

git clone --depth 1 https://github.com/ibelick/ui-skills.git
mkdir -p ~/.claude/skills
cp -R ui-skills/skills/improve-ui ~/.claude/skills/

Claude Code notices new or changed skills in these folders during the session. If the skills folder did not exist when the session started, run /reload-skills once. Then type /improve-ui or just describe the task.

On Windows, run the commands in Git Bash or WSL, or copy the folder into the same place with File Explorer.

Codex, Gemini CLI, Cursor and GitHub Copilot

These agents read the same SKILL.md folders. The one folder all four accept inside a project is .agents/skills/, so this copies the skill there:

git clone --depth 1 https://github.com/ibelick/ui-skills.git
mkdir -p .agents/skills
cp -R ui-skills/skills/improve-ui .agents/skills/
AgentProject folderYour user folderNotes
OpenAI Codex
.agents/skills/
~/.agents/skills/
Type $ or run /skills to pick a skill; Codex also chooses one when your task matches.
Gemini CLI
.agents/skills/
.gemini/skills/
~/.agents/skills/
~/.gemini/skills/
Or install from Git with gemini skills install <git URL> --path <folder> --consent. Google replaced Gemini CLI with Antigravity CLI for free-tier and Google One users on 18 June 2026; Antigravity uses .agents/skills/ in a project and ~/.gemini/config/skills/ for you.
Cursor
.agents/skills/
.cursor/skills/
~/.agents/skills/
~/.cursor/skills/
Type / in Agent chat to pick a skill. Cursor also reads Claude Code's .claude/skills/ folders.
GitHub Copilot
.github/skills/
.agents/skills/
.claude/skills/
~/.copilot/skills/
~/.agents/skills/
Works in Copilot CLI, the cloud agent, and agent mode in VS Code, Visual Studio and JetBrains (IDE support depends on the version). Or run gh skill install OWNER/REPOSITORY SKILL (public preview, GitHub CLI 2.90.0 or later).
Only install skills you trust. Anthropic's advice is to use skills only from trusted sources, and to read every file of any other skill before using it: a skill can direct the agent to run commands, read files or send data in ways its description does not mention.

Sources for these steps

Checked against each vendor's documentation on 2026-11-14:

Evidence before redesign

A draft design proposal does not automatically govern current code. The audit must establish which source owns the selected surface.

Plans are separate from implementation and must contain the evidence and exact owners needed by an executor without this conversation.

Questions about improve-ui

Will it edit components?

No. Product source stays read-only; allowed plan files live under design-plans.

Is repetition enough to prove a defect?

No. Each finding needs a binding contract, runtime evidence and one deterministic correction.

Does it always find three problems?

No. Three is a maximum; no supported findings is a valid result.

Does it automatically inspect screenshots?

Only when rendered evidence is supplied or visual inspection is explicitly requested.

More AI skills

Related skills

Skill categories