Skip to content
Back to search

community-access

MarketplaceWorks with Claude CodeGitHub

Accessibility review agents for Claude Code, GitHub Copilot, and Claude Desktop. Eleven specialists that enforce WCAG 2.2 AA compliance so AI coding tools stop generating inaccessible code.

Safety scan

A static text scan, not a guarantee. Select the badge to see what was scanned and every finding.

Checked today

Check it deeper on your machine

It has changed since its scan on 11 Nov 2026.

As a Claude Code plugin, install it and then ask Claude Code:

/plugin marketplace add keithmackay/mackayi/plugin install scanskillsafety@mackayiDo a deep scan of https://github.com/Community-Access/accessibility-agents with scanskillsafety

Or from a terminal, clone it and run it:

git clone https://github.com/keithmackay/scanskillsafety-skill ~/.claude/skills/scanskillsafetynode ~/.claude/skills/scanskillsafety/cli.cjs --deep 'https://github.com/Community-Access/accessibility-agents'

It checks the exact commit: every file, its packages against OSV (api.osv.dev), how it compares with this page's scan (findsafeskills.com receives the repository's address), and, if you agree, Claude reading every file with your own Claude subscription. It never runs the code.

Badge for your README
findsafeskills safety rating

Markdown

[![findsafeskills safety rating](/api/listings/community-access--accessibility-agents/badge.svg)](/listings/community-access--accessibility-agents)

HTML

<a href="/listings/community-access--accessibility-agents"><img src="/api/listings/community-access--accessibility-agents/badge.svg" alt="findsafeskills safety rating" /></a>

Install

Install instructions from the README

As written by the repo's author. findsafeskills has not verified these steps.

As a plugin (on Gemini CLI, an extension), which loads the skills, the hooks and the MCP server registration (on Codex, the skills and the MCP server only: Codex no longer loads plugin hooks, so use `npm run setup` there for the edit gate):

Claude Code   /plugin marketplace add Community-Access/accessibility-agents
              /plugin install accessibility-agents@community-access
Copilot CLI   copilot plugin marketplace add Community-Access/accessibility-agents
              copilot plugin install accessibility-agents@community-access
Codex         codex plugin marketplace add Community-Access/accessibility-agents
              codex plugin add accessibility-agents@community-access
              (skills and MCP server only; no edit gate, see below)
VS Code       Chat: Install Plugin From Source, with the repository URL
              (requires the chat.plugins.enabled setting)
Gemini CLI    gemini extensions install https://github.com/Community-Access/accessibility-agents
              (uses the extension archive attached to each release, from 7.1.0)

The MCP server ships as a single bundle, `mcp-server/dist/stdio.mjs`, so a plugin install starts it with no `npm install`. Only the optional scanners need installing: Playwright and `@axe-core/playwright` for the browser scans, `pdf-lib` for PDF form conversion, and the veraPDF CLI. The tools that need them say how when they are missing.

The MCP server on its own is published to npm as `@community-access/accessibility-agents-mcp` and listed in the MCP Registry as `io.github.Community-Access/accessibility-agents`, from 7.1.0 once the release is published: `npx -y @community-access/accessibility-agents-mcp`. See [the MCP server README](mcp-server/README.md).

Directly from a checkout, for Antigravity, or any client without the plugin:

npm install
npm run setup -- --dry-run
npm run setup

`npm run setup` copies the skills to `~/.agents/skills`, the personal location Codex, the Copilot CLI, VS Code and Gemini CLI read, and to `~/.gemini/config/skills` when Antigravity is present. It writes the hook manifest for Codex, Gemini CLI and Antigravity when they are present. It never overwrites a hooks file or a skill you have edited, and `--dry-run` prints every action without doing it. Claude Code does not read `~/.agents/skills`, so the installer writes nothing for it: use the plugin, or start Claude Code with `--plugin-dir` pointing at the checkout. For the optional scanners, run `npm install --prefix mcp-server`, then `npx playwright install chromium` in `mcp-server/`.

What was scanned

The files the safety scan read, as written by the repository's author and shown as plain text. Lines with a finding are marked.

.claude-plugin/marketplace.json · Manifest · 22 lines

View this file on GitHub (read at the repository's current version: the scan did not record which commit it read)

This file has changed since it was scanned, so this is not exactly the text that was scanned, and no line is marked. The findings are listed with the safety scan above.

  1. {
  2. "name": "community-access",
  3. "owner": {
  4. "name": "Community Access"
  5. },
  6. "metadata": {
  7. "description": "Accessibility enforcement for agentic coding clients"
  8. },
  9. "plugins": [
  10. {
  11. "name": "accessibility-agents",
  12. "source": ".",
  13. "description": "WCAG 2.2 AA enforcement as Agent Skills. Six router skills dispatch a specialist team across web, document, markdown, desktop and GitHub work; hooks block edits to user-facing files until the accessibility lead has reviewed; an MCP server supplies the scanners.",
  14. "version": "7.1.0",
  15. "author": {
  16. "name": "Community Access"
  17. },
  18. "category": "accessibility",
  19. "homepage": "https://github.com/Community-Access/accessibility-agents"
  20. }
  21. ]
  22. }
README.md · README · 208 lines

View this file on GitHub (read at the repository's current version: the scan did not record which commit it read)

This file has changed since it was scanned, so this is not exactly the text that was scanned, and no line is marked. The findings are listed with the safety scan above.

  1. # Accessibility Agents
  2. WCAG 2.2 AA enforcement for agentic coding, as a set of Agent Skills. One
  3. package, read natively by Claude Code, Codex, GitHub Copilot, Gemini CLI,
  4. Antigravity and ChatGPT, with no per-client copies.
  5. Models forget accessibility while generating code. This package makes that
  6. harder: it routes the work to the right specialist, blocks edits to user-facing
  7. files until the accessibility lead has reviewed, and renders audit reports from
  8. structured findings rather than from memory.
  9. ## Install
  10. As a plugin (on Gemini CLI, an extension), which loads the skills, the hooks
  11. and the MCP server registration (on Codex, the skills and the MCP server only: Codex no longer
  12. loads plugin hooks, so use `npm run setup` there for the edit gate):
  13. ```text
  14. Claude Code /plugin marketplace add Community-Access/accessibility-agents
  15. /plugin install accessibility-agents@community-access
  16. Copilot CLI copilot plugin marketplace add Community-Access/accessibility-agents
  17. copilot plugin install accessibility-agents@community-access
  18. Codex codex plugin marketplace add Community-Access/accessibility-agents
  19. codex plugin add accessibility-agents@community-access
  20. (skills and MCP server only; no edit gate, see below)
  21. VS Code Chat: Install Plugin From Source, with the repository URL
  22. (requires the chat.plugins.enabled setting)
  23. Gemini CLI gemini extensions install https://github.com/Community-Access/accessibility-agents
  24. (uses the extension archive attached to each release, from 7.1.0)
  25. ChatGPT (once published) open Plugins in the sidebar, search for
  26. Accessibility Agents, select Install plugin; then type
  27. @accessibility-agents in a chat (skills only, see below)
  28. ```
  29. What it does not do in ChatGPT: the plugin there is the skills alone, from
  30. OpenAI's plugin directory. It does not run the scanners and has no edit gate.
  31. It works from the checker results and document text you paste, and a skill
  32. that would need a scanner asks for your checker's results. See
  33. [the OpenAI section of the plugin standard](docs/standards/agent-plugins.md#openai-plugin-directory-chatgpt-and-codex).
  34. The MCP server ships as a single bundle, `mcp-server/dist/stdio.mjs`, so a
  35. plugin install starts it with no `npm install`. Only the optional scanners
  36. need installing: Playwright and `@axe-core/playwright` for the browser scans,
  37. `pdf-lib` for PDF form conversion, and the veraPDF CLI. The tools that need
  38. them say how when they are missing.
  39. The MCP server on its own is published to npm as
  40. `@community-access/accessibility-agents-mcp` and listed in the MCP Registry as
  41. `io.github.Community-Access/accessibility-agents`, from 7.1.0 once the release
  42. is published: `npx -y @community-access/accessibility-agents-mcp`. See
  43. [the MCP server README](mcp-server/README.md).
  44. Directly from a checkout, for Antigravity, or any client without
  45. the plugin:
  46. ```bash
  47. npm install
  48. npm run setup -- --dry-run
  49. npm run setup
  50. ```
  51. `npm run setup` copies the skills to `~/.agents/skills`, the personal location
  52. Codex, the Copilot CLI, VS Code and Gemini CLI read, and to
  53. `~/.gemini/config/skills` when Antigravity is present. It writes the hook
  54. manifest for Codex, Gemini CLI and Antigravity when they are present. It never overwrites a
  55. hooks file or a skill you have edited, and `--dry-run` prints every action
  56. without doing it. Claude Code does not read `~/.agents/skills`, so the
  57. installer writes nothing for it: use the plugin, or start Claude Code with
  58. `--plugin-dir` pointing at the checkout. For the optional scanners, run
  59. `npm install --prefix mcp-server`, then `npx playwright install chromium` in
  60. `mcp-server/`.
  61. ## How to use it
  62. Describe the problem. Six entry points route the rest.
  63. | You are working on | Skill |
  64. |---|---|
  65. | HTML, JSX, TSX, Vue, Svelte, CSS, or a server template | `accessibility-lead` |
  66. | A full site or app audit rather than one change | `web-accessibility-wizard` |
  67. | Word, Excel, PowerPoint, PDF or ePub files | `document-accessibility-wizard` |
  68. | Markdown documentation | `markdown-a11y-assistant` |
  69. | Python, wxPython, desktop apps, NVDA add-ons, accessibility tooling | `developer-hub` |
  70. | Issues, pull requests, releases, projects, actions, security, teams, wikis | `github-hub` |
  71. Those six dispatch a further 107 specialist, helper and reference skills by
  72. name. The other 107 are invisible to the model until a router names one, which
  73. is why the whole package costs about 1,100 tokens before you have said anything.
  74. ## What it does
  75. - **Reviews web UI** across ARIA, keyboard, focus, forms, contrast, modals, live
  76. regions, headings, tables, links, cognitive load, internationalisation,
  77. performance, mobile, email, data visualisation and web components.
  78. - **Scans documents** for Office, PDF and ePub accessibility, one file or a
  79. whole tree, and generates remediation scripts.
  80. - **Audits markdown** for link text, alt text, heading order, tables, emoji and
  81. diagrams.
  82. - **Blocks unreviewed edits.** A hook refuses a write to a user-facing file
  83. until the accessibility lead completes, on every client that supports hooks.
  84. - **Renders reports.** Findings are JSON; a script writes the report, computes
  85. the score, and shows the delta against the previous run.
  86. - **Supplies scanners over MCP.** Thirty-nine tools: axe-core, Office, PDF,
  87. ePub, markdown, veraPDF, Playwright behavioural scans, audit history and
  88. report rendering.
  89. ## Layout
  90. ```text
  91. AGENTS.md the always-on contract, under 1,200 tokens
  92. CLAUDE.md one line, importing AGENTS.md
  93. plugin.json Agent Plugins 1.0 manifest
  94. mcp.json MCP server registration
  95. skills/ 113 skills
  96. a11y-core/ shared contract, findings schema, scripts
  97. <router>/ the six entry points
  98. <specialist>/ one domain each, dispatched by a router
  99. kb-<domain>/ reference data: rule tables, URL registries, formulas
  100. hooks/ one guard script and the client hook manifests
  101. com.github.copilot/ Copilot and VS Code plugin hooks (Agent Plugins 1.0)
  102. mcp-server/ the scanners
  103. docs/ conformance dossier, guides, release notes
  104. ```
  105. ## Verify
  106. ```bash
  107. npm run verify
  108. ```
  109. Sixteen gates, each naming the standard it checks and what a failure would
  110. cost; all sixteen passed on 2026-10-28. Two are optional: the dependency advisory check, and live sessions on
  111. whichever of Claude Code, Codex and Copilot are installed.
  112. The evidence, measured results and the one documented deviation from the Agent
  113. Skills specification are in
  114. [the conformance dossier](./docs/standards/conformance.md).
  115. ```bash
  116. npm run measure # what the package costs, per client
  117. npm run measure:dispatch # what dispatching a specialist costs
  118. npm run probe:mcp # the MCP tool surface
  119. ```
  120. ## Standards
  121. Each standard, with what it governs.
  122. | Standard | Governs |
  123. |---|---|
  124. | [Agent Skills](https://agentskills.io/specification) | every skill |
  125. | [AGENTS.md](https://agents.md/) | the always-on contract |
  126. | [Agent Plugins 1.0](https://agent-plugins.org/) | packaging |
  127. | [Model Context Protocol](https://modelcontextprotocol.io/) | the scanners |
  128. | [WCAG 2.2](https://www.w3.org/TR/WCAG22/) | every criterion a finding cites |
  129. | [Open Source Accessibility Framework](https://open-source-accessibility.github.io/Open-Source-Accessibility-Framework/) | how this project runs accessibility as an open source project, and what the agents help your project adopt |
  130. ## Privacy and data
  131. Scanning runs on your machine. The package has no telemetry or analytics, needs
  132. no account or key, and does not send your files or results to the project.
  133. Outbound traffic happens only when you ask for it:
  134. - the web scanners load the URLs you give them, in a local browser;
  135. - the five GLOW tools (`glow_health_check`, `glow_audit_document`,
  136. `glow_fix_document`, `glow_convert_document`, `glow_generate_report`) send a
  137. request, and for the document tools the file you name, to the GLOW service at
  138. `https://letitglow.app/mcp`. GLOW is a separate service with its own terms;
  139. use these tools only on files you may share with it.
  140. Details, including what is stored locally, are in [PRIVACY.md](PRIVACY.md).
  141. ## Accessibility of this project
  142. How we work on our own accessibility, how to report a barrier (on GitHub or by
  143. email) and how we respond are in [ACCESSIBILITY.md](ACCESSIBILITY.md). We
  144. follow GitHub's [Open Source Accessibility Framework](https://open-source-accessibility.github.io/Open-Source-Accessibility-Framework/);
  145. the agents can apply the same framework to your repository: ask your client to
  146. adopt the Open Source Accessibility Framework, or one of its four phases
  147. (foundational, workflow, testing, community).
  148. [![Open Source Accessibility Framework progress for Accessibility Agents](https://img.shields.io/badge/Accessibility%20Framework-Accessibility%20Agents-F143AB?logo=github&logoColor=000000)](https://github.com/open-source-accessibility/Open-Source-Accessibility-Framework/issues/52)
  149. Our progress, action by action, is tracked in public in
  150. [our framework tracking issue](https://github.com/open-source-accessibility/Open-Source-Accessibility-Framework/issues/52)
  151. and recorded with its evidence in
  152. [the framework record](docs/accessibility/accessibility-framework.md). We
  153. publish [accessibility progress updates](docs/accessibility/accessibility-progress.md)
  154. each month, and accessibility issues anyone can pick up are labelled
  155. [good first issue](https://github.com/Community-Access/accessibility-agents/issues?q=is%3Aissue+is%3Aopen+label%3Aaccessibility+label%3A%22good+first+issue%22)
  156. and [help wanted](https://github.com/Community-Access/accessibility-agents/issues?q=is%3Aissue+is%3Aopen+label%3Aaccessibility+label%3A%22help+wanted%22).
  157. ## Contributing
  158. Edit `skills/<name>/SKILL.md`, keep the body under 6 KiB, put the long tail in
  159. `references/`, and run `npm run verify`. Details in
  160. [CONTRIBUTING.md](./CONTRIBUTING.md), and the accessibility expectations for
  161. every contribution in [ACCESSIBILITY.md](ACCESSIBILITY.md#contributor-expectations).
  162. ## History
  163. Version 7.0 replaced six hand-maintained per-client trees with this one package.
  164. The plan, the phases and the measured impact are in
  165. [the modernization record](docs/history/2026-09-modernization.md). What comes
  166. next is in [ROADMAP.md](ROADMAP.md).
  167. ## Licence
  168. MIT. See [LICENSE](./LICENSE).

Ratings and comments

Sign in to rate this skill.

Sign in

Free to use. If it’s helped, pass it on.