Getting it into your agent
It runs from inside its repository, so the clone comes first — what it calls does not travel with the file alone.
git clone --depth 1 https://github.com/Community-Access/accessibility-agentsnpx agentmods add skills/community-access/accessibility-agents/web-accessibility-wizardWrote 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/community-access/accessibility-agents/web-accessibility-wizard)<a href="https://agentmods.dev/skills/community-access/accessibility-agents/web-accessibility-wizard"><img src="https://agentmods.dev/badge/skills/community-access/accessibility-agents/web-accessibility-wizard/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/community-access/accessibility-agents/web-accessibility-wizard"><img src="https://agentmods.dev/badge/skills/community-access/accessibility-agents/web-accessibility-wizard.svg" alt="Reviewed on agentmods" width="80" height="20"></a>- NVIDIA SkillSpector warn
SkillSpector: 13 findings, up to high
These are SkillSpector’s own severities. On a checked sample its high-severity flags on skills were ~96% false positives — a documented command, a public API, a “never do X” rule — so we show them as a caution to read, not a verdict. Why →
- high Anti-Refusal · line 795 Skill instructs the agent to omit warnings, disclaimers, or ethical commentary. Stripping safety caveats hides risk from the user and is a common jailbreak preamble.Fix: Remove instructions that suppress warnings, disclaimers, or ethical commentary. Let the agent surface safety-relevant caveats to the user.
- medium Excessive Agency · line 781 Skill enables autonomous high-impact decisions without human-in-the-loop verification. Critical operations (destructive commands, financial transactions, data deletion) should require explicit user confirmation.Fix: Add human-in-the-loop confirmation for destructive, irreversible, or high-impact operations. Never auto-execute commands that modify files, send data, or alter system state.
- medium Excessive Agency · line 1342 Skill enables autonomous high-impact decisions without human-in-the-loop verification. Critical operations (destructive commands, financial transactions, data deletion) should require explicit user confirmation.Fix: Add human-in-the-loop confirmation for destructive, irreversible, or high-impact operations. Never auto-execute commands that modify files, send data, or alter system state.
- medium MCP Rug Pull · line 325 npx commands without a version suffix (e.g. @1.0.0) create a rug-pull risk if the upstream server is compromised and publishes a malicious update.Fix: Pin the version: npx @scope/[email protected]
- medium MCP Rug Pull · line 337 npx commands without a version suffix (e.g. @1.0.0) create a rug-pull risk if the upstream server is compromised and publishes a malicious update.Fix: Pin the version: npx @scope/[email protected]
- medium MCP Rug Pull · line 340 npx commands without a version suffix (e.g. @1.0.0) create a rug-pull risk if the upstream server is compromised and publishes a malicious update.Fix: Pin the version: npx @scope/[email protected]
- medium MCP Rug Pull · line 343 npx commands without a version suffix (e.g. @1.0.0) create a rug-pull risk if the upstream server is compromised and publishes a malicious update.Fix: Pin the version: npx @scope/[email protected]
- medium MCP Rug Pull · line 346 npx commands without a version suffix (e.g. @1.0.0) create a rug-pull risk if the upstream server is compromised and publishes a malicious update.Fix: Pin the version: npx @scope/[email protected]
- medium MCP Rug Pull · line 328 npx commands without a version suffix (e.g. @1.0.0) create a rug-pull risk if the upstream server is compromised and publishes a malicious update.Fix: Pin the version: npx @scope/[email protected]
- medium MCP Rug Pull · line 352 npx commands without a version suffix (e.g. @1.0.0) create a rug-pull risk if the upstream server is compromised and publishes a malicious update.Fix: Pin the version: npx @scope/[email protected]
- medium MCP Rug Pull · line 617 npx commands without a version suffix (e.g. @1.0.0) create a rug-pull risk if the upstream server is compromised and publishes a malicious update.Fix: Pin the version: npx @scope/[email protected]
- medium MCP Rug Pull · line 1489 npx commands without a version suffix (e.g. @1.0.0) create a rug-pull risk if the upstream server is compromised and publishes a malicious update.Fix: Pin the version: npx @scope/[email protected]
- medium MCP Rug Pull · line 619 npx commands without a version suffix (e.g. @1.0.0) create a rug-pull risk if the upstream server is compromised and publishes a malicious update.Fix: Pin the version: npx @scope/[email protected]
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 | $0.00037 | $0.01029 |
| Opus 5.5 | $0.00015 | $0.00412 |
| Sonnet 5.5 | $0.00007 | $0.00206 |
| Haiku 4.5 | $0.00004 | $0.00103 |
Grade A, and why
web-accessibility-wizard 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 1mo 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 — 98 lines — stays where its author put it; the contents beside it link to each section on GitHub.
You run a full, guided accessibility audit of a web application and produce a
scored report someone can act on. You are the long form; accessibility-lead is
the short form for a single change. Web content only: Office and PDF files go to
document-accessibility-wizard, markdown to markdown-a11y-assistant.
Ask before you scan
Do not start scanning. Run Phase 0 first and get answers. An audit aimed at the
wrong pages, the wrong framework or the wrong conformance target wastes the
whole run. references/phase-0-discovery.md has the questions and the
framework-specific follow-ups.
Phases
Read the reference for a phase when you reach it, not before.
| Phase | What happens | Read |
|---|---|---|
| 0 | Discovery: scope, framework, conformance target, tooling present | references/phase-0-discovery.md |
| 1 to 8 | Domain checks, one specialist each, dispatched in parallel | references/phases-1-8-domain-checks.md |
| 9 to 10 | Testing advice, severity scoring, behavioural testing with Playwright | references/phases-9-10-testing.md |
| 11 | Report, scorecard and handoffs | references/phase-11-report-template.md |
| 12 | Continuous integration and scan configuration | references/phase-12-ci-and-config.md |
Screenshot rules and audit scope are in references/screenshots-and-scope.md.
How specialists are dispatched and what context they receive is in
references/delegation-and-scan-context.md.
Dispatching skills
Phases 1 to 8 are one specialist per domain and they are independent, so
dispatch them together rather than in sequence. Pick from
references/dispatch-matrix.md and use exactly this prompt shape. Never read a
skill's instructions into your own context, and never paste its body.
Activate the skill "<skill-name>". If skill activation is unavailable,
read skills/<skill-name>/SKILL.md and follow it.
Task: <one paragraph, what to check and why>
Scope: <file list or glob>
Rules: semantic HTML before ARIA; report, do not edit.
Return ONLY JSON matching skills/a11y-core/schemas/findings.schema.json.
What ships with it
10 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.
- agents/openai.yaml 423 B
- references/delegation-and-scan-context.md 6.2 KB
- references/dispatch-matrix.md 4.3 KB
- references/phase-0-discovery.md 10 KB
- references/phase-11-handoffs.md 11 KB
- references/phase-11-report-template.md 11 KB
- references/phase-12-ci-and-config.md 11 KB
- references/phases-1-8-domain-checks.md 6.6 KB
- references/phases-9-10-testing.md 12 KB
- references/screenshots-and-scope.md 5.3 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.
- 1mo ago Changed · -1,591 lines · -52 tokens per session 1806d64a2e43
- 1mo ago First seen · 1,689 lines · 89 tokens per session scan A e0b98aeb9175
web-accessibility-wizard is a skill published in the GitHub repository Community-Access/accessibility-agents (428 stars, last pushed 1mo ago), licensed MIT. It adds 37 tokens to every session and 1,029 once invoked, about $0.0001 per session on Opus 5.5. 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 2026-09-03.
Other skills, from other repositories
accessibility
Use when making a web UI conform to WCAG 2.2 Level AA — axe-core or Lighthouse a11y violations, keyboard operability, focus management, ARIA roles/names/live regions, contrast, tap-target size. NOT palette or visual intent (that is design), NOT test-runner setup (that is testing-web), NOT LCP/page-speed (that is…
Accessibility Auditor
Comprehensive WCAG 2.1 AA compliance testing combining automated axe-core scans with manual keyboard navigation, screen reader compatibility, and focus management verification.
Accessibility A11y Enhanced
Comprehensive WCAG compliance and accessibility testing covering ARIA, keyboard navigation, screen readers, color contrast, and automated a11y validation.
accessibility-expert
Build interfaces usable by everyone: WCAG 2.2 conformance, semantic HTML, ARIA, keyboard navigation, screen readers and accessible forms. Use when the user mentions accessibility, a11y, WCAG, ARIA, screen readers, keyboard navigation, colour contrast, focus management, the European Accessibility Act or Section 508, or…
monochrome
Build accessible UI with monochrome, the framework-agnostic UI library for agents. Use whenever a project depends on monochrome or has ids starting with mct:, mcc:, or mcr:, and whenever you write or fix an accordion, collapsible, dialog or modal, dropdown or context menu, menubar, popover, tabs, or tooltip in HTML…
accessibility-engineer
You are the Accessibility Engineering Specialist. You ensure digital products are usable by everyone, including people with visual, auditory, motor, and cognitive disabilities. You audit against WCAG 2.2 standards (AA minimum, AAA preferred), implement ARIA patterns, ensure keyboard navigability, test with screen…