Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx skills add magnus919/agent-skills --skill web-accessibilitygit clone --depth 1 https://github.com/magnus919/agent-skillsWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/magnus919/agent-skills/web-accessibility)<a href="https://agentmods.dev/skills/magnus919/agent-skills/web-accessibility"><img src="https://agentmods.dev/badge/skills/magnus919/agent-skills/web-accessibility/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/magnus919/agent-skills/web-accessibility"><img src="https://agentmods.dev/badge/skills/magnus919/agent-skills/web-accessibility.svg" alt="Reviewed on agentmods" width="80" height="20"></a>- NVIDIA SkillSpector pass
What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00095 | $0.01174 |
| Opus 5 | $0.00048 | $0.00587 |
| Sonnet 5 | $0.00019 | $0.00235 |
| Haiku 4.5 | $0.00010 | $0.00117 |
Grade A, and why
web-accessibility scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 4d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 63 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Web Accessibility
Accessibility is successful task completion and recovery for disabled people, not a tool score. Define observable requirements before implementation, prefer native platform behavior where it fits, and gather bounded evidence in the supported browser and assistive-technology (AT) environment.
Workflow
- Define the task, users, supported browsers/AT, input methods, and failure recovery. Start with
assets/acceptance-criteria-template.md. - Choose native HTML first. Load
references/semantics-and-names.mdfor structure, native controls, accessible names, descriptions, and dynamic updates. - Specify every interactive state: trigger, role/name/state/value, keyboard commands, focus entry/exit/return, pointer alternative, errors, and visible feedback. Load the focused pattern reference.
- Design for visual and input adaptation using
references/visual-input-and-motion.md; distinguish WCAG requirements from product goals and exceptions. - Implement only after the contract is defined. For custom composite widgets, first evaluate a proven native or maintained component in the chosen stack; APG examples are informative guidance, not production proof.
- Verify with
references/testing-and-evidence.mdand the design, implementation, and release templates. Record tool, version, scope, manual tasks, actual browser/AT combinations, failures, and blocked or not-applicable items. - Do not issue a WCAG conformance or general accessibility verdict from an automated scan. A conformance claim needs a complete, scoped evaluation outside this skill's automatic verdict.
Load On Demand
| Need | Load |
|---|---|
| Standards status, source authority, or exact links | references/source-index.md |
| Native semantics, labels, accessible names, landmarks, live updates | references/semantics-and-names.md |
| Tab order, focus visibility, traps, restoration, route changes | references/keyboard-focus-and-routing.md |
| Labels, validation, recoverable errors, authentication | references/forms-errors-authentication.md |
| Modal dialogs, disclosure, responsive navigation | references/dialogs-disclosures-navigation.md |
| Tabs, comboboxes, listboxes, grids, menus, other composites | references/composite-widgets.md |
| Contrast, reflow, zoom, targets, drag, motion, media | references/visual-input-and-motion.md |
| Automated checks, manual protocols, AT matrix, evidence | references/testing-and-evidence.md |
| Framework or skill boundary | references/routing.md |
| Validate common outcomes without claiming conformance | references/outcome-probes.md |
What ships with it
16 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
- assets/acceptance-criteria-template.md 1.0 KB
- assets/review-checklist-design.md 755 B
- assets/review-checklist-implementation.md 929 B
- assets/review-checklist-release.md 776 B
- evals/evals.json 2.9 KB
- README.md 2.3 KB
- references/composite-widgets.md 2.4 KB
- references/dialogs-disclosures-navigation.md 1.9 KB
- references/forms-errors-authentication.md 2.3 KB
- references/keyboard-focus-and-routing.md 1.4 KB
- references/outcome-probes.md 1.6 KB
- references/routing.md 1.2 KB
- references/semantics-and-names.md 2.6 KB
- references/source-index.md 2.5 KB
- references/testing-and-evidence.md 3.2 KB
- references/visual-input-and-motion.md 2.6 KB
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 4d ago First seen · 63 lines · 95 tokens per session scan A ae7666592713
web-accessibility is a skill published in the GitHub repository magnus919/agent-skills (75 stars, last pushed today), licensed MIT. It adds 95 tokens to every session and 1,174 once invoked, about $0.0005 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-09-05.
Other skills, from other repositories
gpt-tasteskill
Use this skill for premium frontend interfaces that require GSAP scroll animations, mathematically verified bento grids, and award-quality design. The most demanding taste variant.
redesign-skill
Use this skill when the task involves improving an existing website or application's visual quality. Works within the existing tech stack rather than rewriting from scratch.
brutalist-skill
Use this skill when the task calls for raw, mechanical, data-heavy interfaces with terminal aesthetics. Synthesizes Swiss typographic modernism with aerospace/military terminal design.
minimalist-skill
Use this skill when the task calls for a refined, typographic-first, Notion/Linear-inspired interface. Produces clean, data-dense UIs with strong visual hierarchy through typography and whitespace alone.
stitch-skill
Use this skill when generating DESIGN.md files for Google Stitch, or when creating agent-friendly design specifications that will drive multi-step UI generation. Translates premium design principles into semantic design language.
taste-skill
Use this skill when the task involves building or improving any web UI, frontend component, dashboard, or landing page. Enforces high-end design standards and eliminates generic AI-generated interface patterns.