gate — Claude Code skill: what it does, how to install
Skill for Claude Code runs a UI quality gate before marking tasks complete. You trigger the command manually, preventing the model from calling it autonomously.
MIT ★ 3,216 updated 2026-11-10
/plugin marketplace add plugin87/ux-ui-agent-skills
/plugin install ux-ui-agent-skills@ux-ui-agent-skillsWhat gate does
This skill executes accuracy_report.mjs across components to validate design tokens, WCAG contrast ratios, hardcoded styles, and theme specifications in light and dark modes. It prints the exact N/N pass score and requires resolving all broken checks before closing the review.
The skill depends on the UX/UI Agent Skills repository including the tokens and scripts directories. Installing only the skill folder causes a KIT_MISSING error and halts execution. Automated gates cover objective compliance and accessibility, while visual aesthetics still require human review.
What it helps with
- validate design tokens and WCAG contrast compliance
- detect hardcoded styling values inside UI components
- test responsive layouts and component state rendering
- generate an exact N/N quality report before review
Example prompts for the agent
- Run the gate check and show the accuracy report
- Verify responsive layouts for index.html using verify_responsive.mjs
- Execute gate and resolve all failing accessibility checks
How to install gate
1Plugin marketplace
/plugin marketplace add plugin87/ux-ui-agent-skills/plugin install ux-ui-agent-skills@ux-ui-agent-skillsThis installs the whole ux-ui-agent-skills plugin with all its skills
2Skills CLI (Vercel)
npx skills add plugin87/ux-ui-agent-skills --skill gate -a claude-code3Manually: put the files in place
~/.claude/skills/gate/SKILL.mdCopy the whole skill folder; for one project use .claude/skills/gate/
1Skills CLI (Vercel)
npx skills add plugin87/ux-ui-agent-skills --skill gate -a codex2Manually: put the files in place
.agents/skills/gate/SKILL.mdFor all projects use ~/.agents/skills/
1Skills CLI (Vercel)
npx skills add plugin87/ux-ui-agent-skills --skill gate -a cursor2Manually: put the files in place
.cursor/skills/gate/SKILL.mdCursor also reads .agents/skills/ and .claude/skills/
1GitHub CLI
gh skill install plugin87/ux-ui-agent-skills .claude/skills/gate2Manually: put the files in place
.github/skills/gate/SKILL.md1Skills CLI (Vercel)
npx skills add plugin87/ux-ui-agent-skills --skill gate -a gemini-cli2Platform CLI
gemini skills install https://github.com/plugin87/ux-ui-agent-skills --path .claude/skills/gateManually: put the files in place
.devin/skills/gate/SKILL.mdThe legacy .windsurf/skills/ path still works
Claude.ai
Zip the folder and upload it in Settings → Capabilities → Skills
Step by step, with commands for each platform: How to install skills in Claude Code: five ways
Questions about gate
How to install gate in Claude Code?
Add the marketplace repo by running /plugin marketplace add plugin87/ux-ui-agent-skills, then run /plugin install ux-ui-agent-skills@ux-ui-agent-skills.
What does the gate skill do?
It runs automated test scripts verifying design tokens, WCAG contrast, and styling rules, returning an exact N/N pass count.
What is required to run gate?
It requires Node.js and the full UX/UI Agent Skills repository initialized with npx [email protected] init or the Claude Code plugin. Without the tokens folder, it aborts with a KIT_MISSING error.
Source text
SKILL.md · MIT
--- name: gate description: Run the one-command quality gate and report the real N/N result. Use before claiming any build/review is done. # A command the user runs. Converted from .claude/commands/ so it has a # CLAUDE_SKILL_DIR of its own and its paths resolve on every install route. # disable-model-invocation keeps the behaviour a command had: the user # starts it, the model never does, and its description stays out of the # listing budget the model's skill choice is drawn from. disable-model-invocation: true --- ## Requirements This skill needs the full UX/UI Agent Skills kit: the gates, the references and the design-system library. Install it with `npx [email protected] init` or as the Claude Code plugin. `npx [email protected] add` copies the skill folder and nothing else, and this skill has no guidance-only mode: its whole job is to run the gates and report their real N/N, so without the kit there is nothing for it to do. > **Step 0: is the kit here?** Check once: > `ls ${CLAUDE_SKILL_DIR}/../../../tokens >/dev/null 2>&1 && echo KIT_OK || echo KIT_MISSING` > On `KIT_MISSING`, say plainly that only the skill folder was installed, name > the install command above, and stop. Do not guess the contents of a file you > could not open, and do not report a result you did not produce. Run the full quality gate and report the ACTUAL output, never a remembered or reasoned number. 1. Run: `node ${CLAUDE_SKILL_DIR}/../../../scripts/accuracy_report.mjs` (tokens + contrast + spec + no-hardcode + theme-refs + no-emoji + real-render WCAG + state-aware, light and dark; all-or-nothing). 2. Report the exact `N/N` line it prints. 3. If it is not fully green: - List each failing check verbatim. - Fix the cause (do not suppress the check). - Re-run until green. Do not announce success between failures. 4. For any rendered HTML touched this session, also run the pixel-truth checks the gate cannot prove: - `node ${CLAUDE_SKILL_DIR}/../../../scripts/gate.mjs verify_states.mjs <file> [--dark]` - `node ${CLAUDE_SKILL_DIR}/../../../scripts/gate.mjs verify_responsive.mjs <file>` - RENDER-AND-LOOK: screenshot the harness, park the pointer off it, inspect every state, click each control to confirm the state changed. Honest scope: these gates prove objective correctness (tokens, a11y, no drift). They do not prove taste. Pair with `DS_REQUIRE_BROWSER=1 node ${CLAUDE_SKILL_DIR}/../../../scripts/taste_audit.mjs` + human review; never claim auto-100% on aesthetics.
© Thientan Soparat · MIT · plugin87/ux-ui-agent-skills