Accessibility Skill
> Adds automated accessibility (a11y) testing to test suites on TestMu AI cloud by enabling WCAG scans through driver capabilities. Framework-agnostic, works with Selenium, Playwright, and Cypress. Use when user mentions "accessibility", "accessibility testing", "a11y scan", "WCAG compliance", "accessibility audit LambdaTest", "is my page accessible".
Install
one command, takes just this skill from the repositorynpx skills add https://github.com/LambdaTest/agent-skills --skill accessibility-skillThe instruction itself
11 sections, as written by the authorAccessibility Testing Skill
Accessibility testing on TestMu AI is layered onto an existing automation suite. You do not
install a separate tool: you enable a capability on your driver and either trigger scans at
chosen points or let every navigation scan automatically. Results appear on the Accessibility
dashboard.
Know the one hard limit up front: the automation capabilities can *start* a scan, but there
is no results API, no exit code, and no build-gating flag. To read a report back programmatically
(for example, to fail a build), use the Accessibility MCP server (see Core Patterns). Automated
scanning also catches only a portion of WCAG; it never replaces manual and screen-reader testing.
Core Patterns
Enable accessibility (capabilities)
// One master switch is required; everything else is optional refinement.
capability.setCapability("accessibility", true); // required
capability.setCapability("accessibility.wcagVersion", "wcag21aa"); // wcag21a | wcag21aa
capability.setCapability("accessibility.bestPractice", false); // default false
capability.setCapability("accessibility.needsReview", true); // include ambiguous issues
Selenium + accessibility
// Option A: trigger a scan at a chosen point (recommended - precise, faster)
capability.setCapability("accessibility", true);
driver.get("https://example.com");
driver.executeScript("lambda-accessibility-scan"); // scan the current page; call as often as needed
// Option B: auto-scan every navigation (Selenium only)
capability.setCapability("accessibility", true);
capability.setCapability("accessibility.autoscan", true);
Run with your normal command, e.g. mvn test. If you use the hook and never call it, no report
is generated.
Playwright + accessibility
const capabilities = {
"accessibility": true,
"accessibility.wcagVersion": "wcag21a",
"accessibility.bestPractice": false,
"accessibility.needsReview": true
};
// Playwright needs the internal scan extension loaded once, after page creation.
// Use Chrome (not Chromium). There is no autoscan/hook on this path.
await ltPage.goto("chrome://extensions/?id=johgkfjmgfeapgnbkmfkfkaholjbcnah");
const toggle = ltPage.locator('#crToggle').nth(0);
await toggle.click();
Run with npx playwright test.
Cypress + accessibility
// lambdatest-config.json - capabilities are dotted here
{ "accessibility": true, "accessibility.wcagVersion": "wcag21aa", "accessibility.needsReview": true }
Register the scanner (file differs by Cypress version):
- v10+: import in
cypress/support/e2e.js, plugin incypress.config.js - v9: import in
cypress/support/index.js, plugin incypress/plugins/index.js
Run with lambdatest-cypress-cli run.
Native mobile app (Appium)
# The scan hook is mandatory for native apps (no autoscan mode).
caps["accessibility"] = True
driver.execute_script("lambda-accessibility-scan")
Read results programmatically (MCP server)
The automation paths write to the dashboard only. The Accessibility MCP server is the one
path that returns a report to the caller:
getAccessibilityReport- scan a public URL and return the reportbuildLocalAppForAnalysis- build and serve a local app, then scan it (before deploy)AnalyseAppViaTunnel- scan a local app already running behind a tunnel
Use it for a real "scan after every build" loop where an agent must act on the result.
Authentication
export LT_USERNAME="your-username" # from your TestMu AI Profile page
export LT_ACCESS_KEY="your-access-key"
Quick Reference
| Task | How |
|------|-----|
| Enable scanning | accessibility: true capability (required) |
| WCAG level | accessibility.wcagVersion = wcag21a or wcag21aa |
| Best-practice checks | accessibility.bestPractice = true (default false) |
| Ambiguous issues | accessibility.needsReview = true |
| Scan at a point | driver.executeScript("lambda-accessibility-scan") |
| Scan every navigation | accessibility.autoscan = true (Selenium only) |
| Read results in code | Accessibility MCP server (dashboard-only otherwise) |
| Dashboard | https://accounts.lambdatest.com/dashboard -> Accessibility |
Deep Patterns
For per-framework setup, the full capability reference, the accessibility score, severity
levels, the WCAG rule-to-fix mapping (remediation), and what automation does and does not cover,
see reference/playbook.md.
Other skills for the same job
different authors, same section of the catalogue
Analyze product screenshots to extract feature lists and generate development task checklists. Use when: (1) Analyzing competitor product screenshots for feature extraction, (2) Generating PRD/task lists from UI designs, (3) Batch analyzing multiple app screens, (4) Conducting competitive analysis from visual references.
| Shared workflow for browser-based review of user-visible frontend changes in Langfuse. Use when a change affects UI behavior, layout, styling, navigation, or browser-visible regressions and should be checked with the Playwright MCP server before signoff.
Optimize web performance through network protocols, resource loading, and browser rendering internals. Use when the user mentions "my site is slow", "Core Web Vitals", "HTTP/2 or HTTP/3", "resource hints", "network latency", "render blocking", "TCP/TLS optimization", "service worker", "Cache-Control or caching strategy", or "critical rendering path". Also trigger when diagnosing slow page loads, optimizing time to first byte, choosing between WebSocket and SSE, or reducing bundle sizes. For UI visual performance, see refactoring-ui. For font loading, see web-typography.
| 将任意文本、网页或 URL 生成符合 Claude/Anthropic 设计语言的 HTML 信息卡片，通过 Playwright 截图为 PNG。 支持 14 种格式：平台封面（公众号、视频号、B站、抖音）、图文内容卡（小红书、教程、对比分析）、 社交分享卡（金句、数据、方形）、长文编辑排版（Broadsheet、Feature、Reader、Digest）。 当用户提到「信息卡、卡片、封面、图文笔记、排版、截图、生成图、内容卡」时使用本技能。
Accessibility testing toolkit using Selenium WebDriver 4+ with Java 21+ and axe-core engine. Use when asked to validate WCAG 2.2 AA compliance, scan pages or components for a11y violations, test keyboard navigation, audit color contrast, check ARIA semantics, generate accessibility reports, filter axe rules, debug screen reader issues, or implement POUR principles (perceivable, operable, understandable, robust).
Run browser inspection for websites and apps. Use after building or changing frontend work to open the rendered site, inspect console and network errors, verify layout and interactions, test links/forms/media, capture screenshots, check responsive behavior, and perform the second visual polish pass before final delivery.
Optimize web performance through network protocols, resource loading, and browser rendering internals. Use when the user mentions "page load speed", "Core Web Vitals", "HTTP/2", "resource hints", "network latency", or "render blocking". Covers TCP/TLS optimization, caching strategies, WebSocket/SSE, and protocol selection. For UI visual performance, see refactoring-ui. For font loading, see web-typography.
Create and refactor HTML5/JavaScript web designs from specifications or descriptions. Generates complete, accessible, responsive web designs with modern frameworks. Automatically verifies designs using Playwright MCP for accessibility and functionality testing. Use this skill when users ask to create web designs, mockups, landing pages, web applications, or refactor existing HTML/CSS/JS designs.
How to use it
Take lambdatest/accessibility-skill from the repository into ~/.claude/skills for personal
use, or into .claude/skills inside a project.
The agent identifies a skill by the name field in its header. Two skills with the
same name cannot sit side by side — one of them will be ignored.
The instructions reference npx.
Without those the skill loads but fails at the first command.