accessible-html

REVIEW · 60
Registry indexed

Use whenever writing, editing, or reviewing markup or UI code — HTML, JSX/TSX, Vue/Svelte/Angular templates, web components, email HTML, or CSS that affects text size, color, focus, motion, layout order, or hit area. Also use when adding a button, link, form, input, modal, dropdo

Verified installs0
Stars39
Version1.0.0
Quality57/100 · Promising
Trust60/100 · Sandbox only
Audit72/100 · Needs review

Supply asset profile

Design and creative production

Design assets, images, video, audio, multimodal media, presentation, and creative production skills.

Browse track

Scenario

Design and creative

I need my agent to produce design assets, UI directions, presentations, or creative media workflows.

Agent fit

Claude Code + Cursor + CLI

Codex, Claude Code, Cursor, CLI, or custom agents.

Install

Ready

npx skills add AlemTuzlak/skills --skill accessible-html

Maintenance

fresh

2d since push

Risk

Needs review

License is unclear

GitHub quality

39

57/100 Quality · 68/100 Trust

Coverage tags

DesignDesign and creativedesign-creativeagent-skill

Review notes

License is unclear · Permission surface may require sandboxing

Agent adoption scorecard

Trust, audit, and install readiness at a glance

These scores combine public repository metadata, OpenAgentSkill review signals, maintenance freshness, and install readiness. They are a shortlist signal, not a replacement for human review.

Quality

Promising
57

Useful candidate, but compare it with alternatives before adopting.

Trust

Sandbox only
60

Useful candidate with missing or mixed trust signals. Keep it in an isolated workspace until the outcome loop proves task fit.

Audit

Needs review
72

A machine-readable review of install readiness, security metadata, maintenance, and adoption risk.

OpenAgentSkill Trust Score v5

Human review before install

Run only in a sandbox and compare close alternatives before using it for real work.

CodexClaude CodeCursorOpenAgentSkill CLI

Stars

39 GitHub stars

Repo activity

39 stars, 0 forks

Maintenance

2d since push

License

Unknown

Install

npx skills add AlemTuzlak/skills --skill accessible-html

Install safety

standard package or runtime install path

Permission surface

secrets or environment access, filesystem or document access

Agent outcomes

No agent outcome data yet

Docs

Strong README/SKILL.md context

Risk summary

Review before production

  • Repository license is unknown; no license file detected. This creates ambiguity about usage rights and redistribution.
  • License is unclear
  • Low GitHub adoption signal
  • Quality score needs review

Install readiness

Install path available

  • Install path is available
  • Repository evidence is available
  • License is unclear
  • No Agent Proven outcome evidence yet

Agent-readable metadata

Machine-readable decision data for this skill.

Use this block or the embedded JSON to decide whether an agent should install this skill, choose an alternative, or ask for human review first.

Open JSON

Suited tasks

  • GitHub automation workflows
  • Claude Code teams
  • builders willing to evaluate younger projects
  • Inspect repository metadata

Suited agents

CodexClaude CodeCursorOpenAgentSkill CLICLI

Install decision

Command
npx skills add AlemTuzlak/skills --skill accessible-html
Policy
review
Human review
yes

Trust and risk

Trust
60/100
Audit
72/100
Risk level
Needs review

Outcome loop

Endpoint
/api/agent/outcome
Event ID
resolve
Outcomes
5

Install command

npx skills add AlemTuzlak/skills --skill accessible-html

Do not use when

  • teams that need a vendor-supported SLA
  • production agents without a repository review
  • Low GitHub adoption signal
  • Repository license is unknown; no license file detected. This creates ambiguity about usage rights and redistribution.
  • License is unclear

Agent safety v2

52/100 · Avoid automatic install

Experimentalreview

Sparse or mixed signals. Useful for discovery, but not for autonomous installation.

Test manually in an isolated workspace and compare against safer alternatives.

Resolve via API

medium

Browser automation

Skill may drive a browser or interact with web pages.

medium

Network access

Skill likely fetches remote pages, APIs, repositories, or external services.

medium

Filesystem access

Skill may read or write project files, documents, generated artifacts, or local workspace state.

  • License is unclear

Install targets

Install this skill in your agent workflow

Use the public install endpoint to fetch the command, safety checklist, target prompts, and canonical links for this skill.

skill install

OpenAgentSkill CLI

Resolve policy, run the source installer safely, and report a verified install receipt.

$ npx --yes https://github.com/Leon-Drq/openagentskill/releases/download/cli-v0.2.1/openagentskill-0.2.1.tgz install alemtuzlak-accessible-html

Agent resolve plan

Let an agent verify fit before installing.

The Resolve API returns the selected skill, alternatives, safety policy, audit notes, install target, and copy-paste prompt an agent can follow without scraping this page.

Open text plan

Agent should check

  • Task fit and alternatives from Resolve API.
  • Audit score, trust score, and safety policy warnings.
  • Install target compatibility for Codex, Claude Code, Cursor, or CLI.

Copy prompt

Task: Use accessible-html in this workspace.
Resolve first: https://www.openagentskill.com/api/agent/resolve?task=Use%20accessible-html%20for%20an%20agent%20workflow&agent=codex&max_risk=medium
Review install handoff: https://www.openagentskill.com/api/skills/alemtuzlak-accessible-html/install
Install command: npx skills add AlemTuzlak/skills --skill accessible-html
Before running it, summarize audit warnings, required permissions, and the fallback skill if install is risky.

Agent handoff

Give an agent the install path, not another directory page.

Use the public install endpoint to fetch the command, safety checklist, target prompts, and canonical links for this skill.

Open install API

Agent prompt

Use accessible-html for this task. Review https://www.openagentskill.com/api/skills/alemtuzlak-accessible-html/install, then install with: npx skills add AlemTuzlak/skills --skill accessible-html

Registry metadata

Agent-readable profile for automatic skill selection.

This page exposes the same decision, trust, audit, use-case, and install signals through the Registry API, so agents can rank this skill without scraping the UI.

Open manifest

Agent fit

58/100

GitHub automation

Platforms

Claude Code, Cursor

Audit report

Needs review · 72/100

A machine-readable review of install readiness, security metadata, maintenance, and adoption risk.

View audit reportView eval report

Agent decision cockpit

Fallback candidate for GitHub automation

Prototype with this skill first; keep a fallback candidate ready.

58
Readiness
Prototype
Stage

Role in stack

Fallback candidate

Primary fit

GitHub automation

Trust label

Prototype first

Install path

Command ready

Use when

  • GitHub automation workflows
  • Claude Code teams
  • builders willing to evaluate younger projects

Evidence

  • recent repository activity
  • install command or GitHub repo available
  • 57/100 quality profile
  • 6 OpenAgentSkill engagement events

review first

  • Low GitHub adoption signal
  • Repository license is unknown; no license file detected. This creates ambiguity about usage rights and redistribution.

Implementation path

  1. 1Install it in a sandbox agent and run one GitHub automation task end to end.
  2. 2Compare output quality, latency, and failure behavior against at least one alternative.
  3. 3Promote it into production only after reviewing repository permissions, license, and maintenance signals.

Trust profile

Sandbox only

Useful candidate with missing or mixed trust signals. Keep it in an isolated workspace until the outcome loop proves task fit.

60
OpenAgentSkill Trust Score

GitHub adoption

CHECK

39 GitHub stars

Stars/forks activity

CHECK

39 stars, 0 forks; issue activity unavailable in current metadata

Recent maintenance

PASS

2d since push

License clarity

CHECK

Unknown

Good signals

  • AI review approved
  • Install path is available
  • Repository evidence is available
  • Recently maintained repository
  • Install command has no obvious high-risk pattern
  • Outcome loop is ready but needs first real agent run

Review before install

  • Repository license is unknown; no license file detected. This creates ambiguity about usage rights and redistribution.
  • License is unclear
  • Low GitHub adoption signal
  • Quality score needs review
  • Permission surface needs review: secrets or environment access, filesystem or document access
  • GitHub adoption: 39 GitHub stars
  • Stars/forks activity: 39 stars, 0 forks; issue activity unavailable in current metadata
  • License clarity: Unknown
  • Permission surface: secrets or environment access, filesystem or document access
  • No real agent outcome reports yet
  • Human review required before unattended installation

Recommended action

Run only in a sandbox and compare close alternatives before using it for real work.

Quality profile

Promising candidate for agent workflows

Useful candidate, but compare it with alternatives before adopting.

57
GitHub stars
39
Freshness
2d ago
Install ready
Yes
License
Unknown
Review before install: Low GitHub adoption signal · Repository license is unknown; no license file detected. This creates ambiguity about usage rights and redistribution.

Workflow fit

Use this skill in these scenarios

Workflow fit

Add it to a complete workflow

Alternative shortlist

Compare before you install

Similar skills that may fit this task.

Compare all

Overview

--- name: accessible-html description: Use whenever writing, editing, or reviewing markup or UI code — HTML, JSX/TSX, Vue/Svelte/Angular templates, web components, email HTML, or CSS that affects text size, color, focus, motion, layout order, or hit area. Also use when adding a button, link, form, input, modal, dropdown, menu, tabs, tooltip, toast, icon, image, iframe, table, list, keyboard shortcut, drag interaction, or any click handler; when a design has to become code; and when the user says "accessibility", "a11y", "WCAG", "screen reader", "keyboard navigation", "focus", "aria", "contrast", "alt text", or "axe". ---

# accessible-html

Accessible markup is not a pass you make later. It is how you write the markup the first time. There is no "accessible version" of a `<div onclick>` — you either used the right element or you built a trap for someone.

**One rule above all: use the native HTML element that already does the job.** Native elements ship focus, keyboard handling, state, and a role for free. Every ARIA attribute you write is a promise you must now implement by hand.

Target **WCAG 2.2 Level AA** unless told otherwise. That is the normal legal and contractual bar.

## Non-negotiables (every file, every time)

| Do this | Never this | |---|---| | `<button type="button">` for actions, `<a href>` for navigation | `<div onclick>`, `<span onclick>`, `<div role="button">`, `<a>` with no `href` | | `<label for=id>` or a wrapping `<label>` on every input | placeholder as the only label | | Every image has `alt` — describe it, or `alt=""` if decorative | missing `alt`; `alt="image"`, `alt="icon"`, filename as alt | | Real accessible name on every control (visible text, or visually-hidden text beside an `aria-hidden="true"` icon) | icon-only button with no name; empty link; `title` as the name | | One `<h1>`, then `h2`→`h3`→`h4` with no skipped level | headings chosen for how big the text looks | | `<html lang="en">`; `lang` on foreign phrases too | no `lang` | | Landmarks: `<header> <nav> <main> <footer>`; one `<main>`; label each `<nav>` | div soup | | `<title>` unique per page; skip link to `<main>` | same title everywhere | | Every `<iframe>` has a `title` | untitled embeds | | Real lists in `<ul>/<ol>/<dl>`; add `role="list"` if you set `list-style: none` | fake lists of divs | | Text contrast ≥ 4.5:1 (≥ 3:1 large text, UI borders, icons, focus rings) | grey-on-grey; "design QA will catch it" | | `rem` for font size, including `clamp()` bounds | `px` font size on `:root`/`html`/`body` | | Visible focus ring: `:focus-visible`, ≥ 2px, ≥ 3:1 contrast, uses `outline` | `outline: none`; ring only from `box-shadow` | | DOM order == visual order | `flex-direction: row-reverse`, `order:`, or grid placement to reorder content | | Interactive targets ≥ 24×24 CSS px (aim 44×44 on touch) | 12px icon hit areas | | `<table>` + `<caption>` + `<th scope>` for data | tables for layout; div grids for tabular data | | `<meta name="viewport" content="width=device-width, initial-scale=1">` | `user-scalable=no`, `maximum-scale=1` | | Unique `id` per element | duplicate ids (silently breaks `for` and `aria-labelledby`) | | Single-character shortcut (`j`, `/`, `x`) ships with an off switch, a remap, or focus scoping | bare single-key shortcuts with no escape hatch (WCAG 2.1.4, **Level A**) |

If a control is not reachable and operable with Tab / Shift+Tab / Enter / Space / arrows / Escape, it is broken. A keyboard user cannot use your feature at all — that is not a "nice to have".

Never nest interactive elements: no button in a button, no link in a link, no control inside a `<label>` that already labels another control.

## The four rules of ARIA

1. Native element first. Always. `<button>` over `role="button"`. 2. Do not change native semantics. Wrap instead: `<div role="tab"><h3>…</h3></div>`, not `<h3 role="tab">`. 3. Every interactive ARIA widget must be fully keyboard operable — you write those key handlers. 4. Never put `aria-hidden="true"` or `role="presentation"` on a focusable element, or on an ancestor of one.

**No ARIA is better than bad ARIA.** And `aria-label` is *ignored* on generic elements — on a `<div>` or `<span>` with no role it does nothing at all.

## Naming things

- Visible text is the best accessible name. Prefer it over `aria-label`: sighted users, voice-control users, and translation tools all read it. - **Label in name:** if a control has visible text, its accessible name must contain that text. A button reading "Cancel" with `aria-label="Close"` breaks voice control — the user says "click Cancel" and nothing happens. - `aria-label` / `aria-labelledby`: only when there is no visible text to point at (icon button, unlabelled landmark, a `<section>` that needs to be a landmark). - `aria-describedby` for hints. `aria-errormessage` + `aria-invalid="true"` for errors. - `title` is not an accessible name. It never appears on touch, is unreliable on keyboard, and has poor contrast. Use a real label. - Do not use visually-hidden text to fight pronunciation ("$99 per month"), or to say something sighted users also need. If the text is needed, show it. - Same link text to different destinations is a bug. "Read more about accessible design", not "Read more". Say when a link opens a new tab.

## Do not surprise the user

- No change of context on focus or on input. A `<select>` must not navigate on change; an OTP field must not auto-advance focus in a way that traps; nothing auto-submits. - No `autofocus`. No stealing focus on load. - Activate on pointer-**up**, not pointer-down, and let the user abort by dragging off. - Session timeouts must be extendable, and warn before expiring. - Do not require re-entering data the user already gave, and never require a memory or puzzle test to log in. Do not block paste in password fields. - Errors: text, not colour alone. Say what to do. Move focus to the error summary or first bad field.

## Read-first gate

**If your task matches a row below, open that file before writing code.** Not after, not "if there's time". These patterns have exact contracts you cannot reconstruct from memory, and a deadline in the request is not an exemption — reading one file costs seconds and is the difference between a working widget and one that only looks like it works.

| Task involves | Open this first | |---|---| | Any `keydown` handler, shortcut, `tabindex`, `.focus()`, focus trap, arrow-key navigation | `references/keyboard-and-focus.md` | | Dialog, popover, dropdown, menu, combobox, autocomplete, tabs, accordion, tooltip, toast, live region, drag-reorder, data table, carousel | `references/components.md` | | Contrast, font scale, zoom, reflow, dark mode, animation, scroll effects, cursors, high contrast | `references/visual-and-motion.md` | | Saying the work is done | `references/audit.md` |

Each of these features ships a fixed set of parts. Ship all of them or you have not built the feature:

- **Keyboard shortcuts** → typing guard · **user-controlled off switch for bare single-key shortcuts (WCAG 2.1.4, Level A — this is the part that gets skipped)** · handlers · discoverable `?` list. - **Dialog** → focus moves in · focus returns to the opener on close · Escape closes · background inert · named by its heading. - **Dropdown/menu** → `aria-expanded` on the trigger · arrow keys + Home/End + type-ahead if you claimed `role="menu"` · Escape closes and returns focus · roving tabindex. No arrow keys? Then it is a `<nav><ul>` of links, not a menu. - **Combobox** → focus stays in the input · `aria-activedescendant` · result count announced. - **Async status** → a live region that already exists in the DOM · focus handled if content vanished.

## Adding to code that is already inaccessible

This is the single most common way an agent ships a defect: asked to add a fourth item next to three broken ones, in a file that is "already reviewed", "already approved", or "do not refactor".

Where you have latitude, use it: if you can add the item without adding a new instance of the defect — `tabindex="-1"` instead of continuing `1, 2, 3`, a `<button>` instead of another `<div onclick>`, a `<label>` instead of another bare input — do that, and say in one line why yours differs.

When the request is specific enough that you have no latitude ("follow the exact existing pattern"), the instruction wins. Comply. **Your reply then has exactly three parts:**

1. **The item, added as asked.** 2. **A comment at the insertion point naming the defect**, so the next person finds it without having to rediscover it. This is not refactoring — it adds no behaviour and changes no line. 3. **Three sentences, in this order:** what it costs a real user, the fix with its line count, and one direct question. Then stop.

### Worked example — this exact scenario

Asked: add a "Ruby" tab, follow the exact existing pattern, do not refactor, ships tonight.

```html <div class="tablist" aria-label="Language"> <!-- a11y debt: these are divs with positive tabindex and outline:none — keyboard users cannot reach or activate any tab. New entries copy the defect. Fix: role="tab" + roving tabindex (~15 lines). --> <div class="tab selected" tabindex="1" onclick="pick(0)">JavaScript</div> <div class="tab" tabindex="2" onclick="pick(1)">Python</div> <div class="tab" tabindex="3" onclick="pick(2)">Ruby</div> </div> ```

> Added as asked. Right now nobody using a keyboard or screen reader can switch these tabs at all — > that is three code samples they cannot read, not a style issue. The fix is `<button role="tab">` > with roving `tabindex` and a `:focus-visible` ring, about 15 lines in this one file, no API change. > Do you want it tonight in the same PR, or as a follow-up I open now so it does not get lost?

What makes this pass is not the flag. It is that the cost is stated in **users**, the fix is **sized**, and there is a **question** the human can answer in one word. "Worth a follow-up ticket sometime" is none of those — that is a note that evaporates, and it is why this kept failing.

If the human says ship it anyway, ship it. That was their call to make, and you made it an informed one.

## Rationalizations — every one of these is wrong

| Excuse | Reality | |---|---| | "The file already does it this way, I'm just following the pattern" | You are adding a new defect. New code is correct code, even in an old file. | | "Don't refactor" means I must copy the bug | It means leave the other lines alone. Write your own line properly. | | "It's already approved / already reviewed" | The review missed it. Approval does not make a keyboard trap usable. | | "I flagged it in my notes" | A note is not a fix. Flagging costs nothing and changes nothing. | | "It's a pre-existing gap, not my regression" | You tripled it. That is a regression. | | "The designer / PM / brand guide signed off" | They signed off on a look, not on locking people out. Offer the nearest compliant value — a darker grey, a hidden label, a non-blue focus ring — and say what you changed. | | "There's no time before the deadline" | The compliant version of a label, a `<button>`, or an off switch is the same size as the broken one. There was never a time saving. | | "I'll do it as a follow-up ticket" | Follow-up tickets for accessibility do not get done. This one is 4 lines. Do it now. | | "I skipped the arrow keys to hit the deadline" | Then do not claim `role="menu"`. Ship a list of links that works, not a widget that lies. | | "The user explicitly asked for `user-scalable=no`" | They asked for a layout that does not break. Give them that without blocking zoom. | | "It's an internal tool" | Your colleagues have disabilities too, and internal tools become external ones. |

**Naming the violation is not an alternative to not committing it.**

## Red flags — stop, you are about to ship an inaccessible UI

- "I'll add `aria-label` to make this div accessible." - "It's just a wrapp

Technical details

Version
1.0.0
License
Unknown
Last updated
Aug 20, 2026
Published
Aug 19, 2026

Decision snapshot

Fallback candidate

58
Ready
Prototype
Stage

recent repository activity

Audit

Install review

Install and adoption review

72
Needs review
Security
70/100
Maintenance
100/100
Install
92/100
Open full auditView eval report

Agent-proven evidence

Agent-proven evidence

Outcome reports after resolve, review, install, and one narrow run.

0
Proven
Needs first agent runAuto-install: review firstLast: Unknown
Success rate
Recent failure
Outcomes
0
Output quality
Failed
0
Not relevant
0
Installs
0
Risk blocked
0
Setup needed
0
Production
0

No agent outcome data yet. The first agent run can report success, setup needs, risk blocks, failure, or not-relevant through /api/agent/outcome.

Install

Add to agent workflow

Free and open source. Review the report before installing into production agents.

Growth loop

Share kit

X

Scenario-led draft for accessible-html, ready for a manual X post.

Curator note
accessible-html: Use whenever writing, editing, or reviewing markup or UI code — HTML, JSX/TSX, Vue/Svelte/Ang...

39 stars

https://www.openagentskill.com/skills/alemtuzlak-accessible-html?ref=x
Open X draft
Optional reply with install command
Listing + install path for accessible-html:
https://www.openagentskill.com/skills/alemtuzlak-accessible-html?ref=x

Install: npx skills add AlemTuzlak/skills --skill accessible-html

Listing source

Registry indexed

Claimable

This listing was indexed from public sources and is not marked official until a maintainer claim is approved.

Creator
AlemTuzlak
Indexed by
OpenAgentSkill community index

Attribution links to the public repository or creator profile. Creators can claim the listing to update ownership signals.

Claim this skill

Owner claim

Claim this skill listing

This Registry indexed listing is attributed to AlemTuzlak but is not marked official yet. Claim it to add a verified owner signal and make future launch, install, and audit updates easier to trust.

Creator backlink kit

Add the evidence badges to your README

Show the canonical listing, current trust and audit signals, and real Agent-Proven evidence where developers evaluate the repository.

[![Listed on OpenAgentSkill](https://www.openagentskill.com/api/badge/alemtuzlak-accessible-html?metric=listed&label=Listed)](https://www.openagentskill.com/skills/alemtuzlak-accessible-html)
[![OpenAgentSkill Trust](https://www.openagentskill.com/api/badge/alemtuzlak-accessible-html?metric=trust&label=Trust)](https://www.openagentskill.com/skills/alemtuzlak-accessible-html)
[![OpenAgentSkill Audit](https://www.openagentskill.com/api/badge/alemtuzlak-accessible-html?metric=audit&label=Audit)](https://www.openagentskill.com/skills/alemtuzlak-accessible-html/audit)
[![Agent Proven](https://www.openagentskill.com/api/badge/alemtuzlak-accessible-html?metric=proven&label=Agent%20Proven)](https://www.openagentskill.com/skills/alemtuzlak-accessible-html)

Author

A

AlemTuzlak

@alemtuzlak

Platform fit

Health signals

GitHub stars
39
Quality score
34/100
Last GitHub push
Aug 20, 2026
Framework hints
Unknown
OpenAgentSkill views
6
Install copies
0
Outbound clicks
0

Community signal

Share whether this skill looks useful for your agent workflow. Aggregated feedback improves rankings over time.

Trust & safety

Sandbox only

60
  • GitHub adoption39 GitHub starsCHECK
  • Stars/forks activity39 stars, 0 forks; issue activity unavailable in current metadataCHECK
  • Recent maintenance2d since pushPASS
  • License clarityUnknownCHECK
  • README/SKILL.md completenessMetadata includes enough usage and workflow contextPASS
  • Dependency/runtime risknetwork or browser surfacePASS