-
Notifications
You must be signed in to change notification settings - Fork 6
MCP Tools
The A11y Agent Team MCP server provides tools that agents use automatically. These same tools are available in Claude Desktop (via the .mcpb extension), GitHub Copilot (via .vscode/mcp.json), and any MCP-compatible client.
Calculate WCAG contrast ratios between two hex colors. Returns the ratio and pass/fail for normal text (4.5:1), large text (3:1), and UI components (3:1).
Get detailed WCAG AA guidelines for a specific component type: modal, tabs, accordion, combobox, carousel, form, live-region, navigation, or general. Returns requirements, code examples, and common mistakes.
Analyze HTML for heading hierarchy issues: skipped levels, multiple H1 tags, empty headings, and heading order problems. Returns WCAG criterion references (1.3.1, 2.4.6).
Scan HTML for link accessibility issues. Detects 17 ambiguous text patterns ("click here", "read more", etc.), URLs used as link text, links opening in new tabs without warning, links to non-HTML resources without file type indication, and repeated identical text pointing to different destinations. WCAG references: 2.4.4, 2.4.9.
Validate form input accessibility. Checks every <input>, <select>, and <textarea> for proper labels (<label for>, aria-label, or aria-labelledby), autocomplete attributes on identity/payment fields, and fieldset/legend on radio/checkbox groups. WCAG references: 1.3.1, 1.3.5, 3.3.2, 4.1.2.
Run axe-core against a live URL and return violations grouped by severity (Critical > Serious > Moderate > Minor). Includes affected HTML elements, WCAG criteria, and fix suggestions. Optionally writes a structured markdown report to a file path.
Prerequisites: npm install -g @axe-core/cli
See axe-core Integration for detailed usage.
Scan a Microsoft Office document (DOCX, XLSX, PPTX) for accessibility issues. Parses the ZIP/XML structure and checks for alt text, headings, tables, language, reading order, and more. Returns findings as SARIF 2.1.0 or human-readable markdown.
Scan a PDF document for accessibility conformance against PDF/UA and the Matterhorn Protocol. Checks tagged structure, metadata, bookmarks, form fields, fonts, and encryption. Returns findings as SARIF or markdown.
Extract metadata from Office or PDF documents including title, author, language, page/slide/sheet count, and creation/modification dates.
Scan multiple documents in a single operation. Accepts a directory path and scans all supported files recursively, applying relevant config files.
Generate a VPAT 2.5 / Accessibility Conformance Report (ACR) template pre-populated with all WCAG 2.2 Level A and AA criteria. Merge in findings from agent reviews to produce a publishable conformance document.
See VPAT Generation for detailed usage.
The tools accept inputs as described in each tool's parameters. Agents call these automatically during reviews:
# Claude Code - agents use tools automatically
/accessibility-lead review index.html
# -> The lead reads the file, passes HTML to check_heading_structure,
# check_link_text, check_form_labels as needed
# Copilot - same behavior
@accessibility-wizard audit the signup page
- Accessibility Lead
- Web Accessibility Wizard
- Document Accessibility Wizard
- Alt Text and Headings
- ARIA Specialist
- Contrast Master
- Forms Specialist
- Keyboard Navigator
- Link Checker
- Live Region Controller
- Modal Specialist
- Tables Data Specialist
- Word Accessibility
- Excel Accessibility
- PowerPoint Accessibility
- PDF Accessibility
- Office Scan Config
- PDF Scan Config
- Testing Coach
- WCAG Guide