Skip to main content
Glama
deenesjakoruzh
bilhasry-deriv

Web Accessibility MCP Server

by bilhasry-deriv
OverviewSchemaSourceRelated ServersDiscussionsAdmin
JavaScript
Local

Servidor MCP de accesibilidad web

insignia de herrería

Un servidor MCP (Protocolo de contexto de modelo) que proporciona capacidades de análisis de accesibilidad web utilizando axe-core y Puppeteer.

Características

  • Analice la accesibilidad web de cualquier URL utilizando axe-core

  • Simular el daltonismo (protanopia, deuteranopia, tritanopia) utilizando matrices de color

  • Informes detallados de violaciones de accesibilidad

  • Compatibilidad con agentes de usuario y selectores personalizados

  • Registro de depuración para la resolución de problemas

  • Comprobaciones de accesibilidad exhaustivas basadas en las pautas WCAG

Related MCP server: Cursor A11y MCP

Prerrequisitos

  • Node.js (v14 o superior)

  • npm

Instalación

Instalación mediante herrería

Para instalar Web Accessibility MCP Server para Claude Desktop automáticamente a través de Smithery :

npx -y @smithery/cli install @bilhasry-deriv/mcp-web-a11y --client claude

Instalación manual

  1. Clonar el repositorio:

git clone [repository-url]
cd mcp-web-a11y

  1. Instalar dependencias:

npm install

  1. Construir el servidor:

npm run build

Configuración

Agregue el servidor a su archivo de configuración de MCP (normalmente ubicado en ~/Library/Application Support/Code/User/globalStorage/saoudrizwan.claude-dev/settings/cline_mcp_settings.json ):

{
  "mcpServers": {
    "web-a11y": {
      "command": "node",
      "args": ["/path/to/mcp-web-a11y/build/index.js"],
      "disabled": false,
      "autoApprove": [],
      "env": {
        "MCP_OUTPUT_DIR": "/path/to/output/directory"
      }
    }
  }
}

Variables de entorno

  • MCP_OUTPUT_DIR : Directorio donde se guardarán las salidas de captura de pantalla

    • Necesario para la herramienta simulate_colorblind

    • Si no se especifica, el valor predeterminado es './output' relativo al directorio de trabajo actual

    • Debe ser una ruta absoluta cuando se configura en la configuración de MCP

Uso

El servidor proporciona dos herramientas: check_accessibility para analizar la accesibilidad web y simulate_colorblind para simular el daltonismo.

Herramienta: check_accessibility

Comprueba la accesibilidad de una URL determinada utilizando axe-core.

Parámetros

  • url (obligatorio): La URL a analizar

  • waitForSelector (opcional): selector CSS que se debe esperar antes del análisis

  • userAgent (opcional): cadena de agente de usuario personalizada para la solicitud

Ejemplo de uso

<use_mcp_tool>
<server_name>mcp-web-a11y</server_name>
<tool_name>check_accessibility</tool_name>
<arguments>
{
  "url": "https://example.com",
  "waitForSelector": ".main-content",
  "userAgent": "Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/120.0.0.0 Safari/537.36"
}
</arguments>
</use_mcp_tool>

Herramienta: simular_daltónico

Simula cómo aparece una página web para usuarios con diferentes tipos de daltonismo mediante transformaciones de matriz de color.

Tipos de daltonismo

La herramienta admite tres tipos de simulación de daltonismo:

  1. Protanopia (ciego al rojo) - Utiliza matriz:

    0.567, 0.433, 0
0.558, 0.442, 0
0, 0.242, 0.758

  2. Deuteranopia (ciego a la luz verde) - Utiliza la matriz:

    0.625, 0.375, 0
0.7, 0.3, 0
0, 0.3, 0.7

  3. Tritanopia (ciego azul) - Utiliza matriz:

    0.95, 0.05, 0
0, 0.433, 0.567
0, 0.475, 0.525

Parámetros

  • url (obligatorio): La URL a capturar

  • type (obligatorio): Tipo de daltonismo a simular ('protanopia', 'deuteranopia' o 'tritanopia')

  • outputPath (opcional): ruta personalizada para la salida de la captura de pantalla

  • userAgent (opcional): cadena de agente de usuario personalizada para la solicitud

Ejemplo de uso

<use_mcp_tool>
<server_name>mcp-web-a11y</server_name>
<tool_name>simulate_colorblind</tool_name>
<arguments>
{
  "url": "https://example.com",
  "type": "deuteranopia",
  "outputPath": "colorblind_simulation.png"
}
</arguments>
</use_mcp_tool>

Formato de respuesta

Respuesta de check_accessibility

{
  "url": "analyzed-url",
  "timestamp": "ISO-timestamp",
  "violations": [
    {
      "impact": "serious|critical|moderate|minor",
      "description": "Description of the violation",
      "help": "Help text explaining the issue",
      "helpUrl": "URL to detailed documentation",
      "nodes": [
        {
          "html": "HTML of the affected element",
          "failureSummary": "Summary of what needs to be fixed"
        }
      ]
    }
  ],
  "passes": 42,
  "inapplicable": 45,
  "incomplete": 3
}

Respuesta de simular_daltónico

{
  "url": "analyzed-url",
  "type": "colorblind-type",
  "outputPath": "path/to/screenshot.png",
  "timestamp": "ISO-timestamp",
  "message": "Screenshot saved with [type] simulation"
}

Manejo de errores

El servidor incluye un manejo integral de errores para escenarios comunes:

  • Errores de red

  • URL no válidas

  • Problemas de tiempo de espera

  • Problemas de resolución de DNS

Las respuestas de error incluirán mensajes detallados para ayudar a diagnosticar el problema.

Desarrollo

Estructura del proyecto

mcp-web-a11y/
├── src/
│   └── index.ts    # Main server implementation
├── build/          # Compiled JavaScript
├── output/         # Generated screenshots
├── package.json    # Project dependencies and scripts
└── tsconfig.json   # TypeScript configuration

Edificio

npm run build

Esto hará lo siguiente:

  1. Compilar TypeScript a JavaScript

  2. Hacer que el archivo de salida sea ejecutable

  3. Coloque los archivos compilados en el directorio build

Depuración

El servidor incluye un registro de depuración detallado que se puede consultar en la salida de la consola. Esto incluye:

  • Solicitudes y respuestas de red

  • Estado de carga de la página

  • Estado de espera del selector

  • Cualquier mensaje de consola de la página analizada

  • Progreso de la simulación de color

Problemas comunes y soluciones

  1. Errores de tiempo de espera

    • Aumente el valor de tiempo de espera en el código

    • Comprobar la conectividad de la red

    • Verificar que la URL sea accesible

  2. Errores de resolución de DNS

    • Verifique que la URL sea correcta

    • Comprobar la conectividad de la red

    • Intente utilizar el subdominio www

  3. Selector no encontrado

    • Verificar que el selector exista en la página

    • Espere a que se cargue el contenido dinámico

    • Verifique la fuente de la página para encontrar el selector correcto

  4. Problemas de simulación de color

    • Asegúrese de que los colores de la página estén especificados en un formato compatible (RGB, RGBA o HEX)

    • Verifique si la página utiliza cambios de color dinámicos (puede requerir tiempo de espera adicional)

    • Verifique que el directorio de salida de la captura de pantalla exista y se pueda escribir

Contribuyendo

  1. Bifurcar el repositorio

  2. Crear una rama de características

  3. Confirme sus cambios

  4. Empujar hacia la rama

  5. Crear una solicitud de extracción

Licencia

Este proyecto está licenciado bajo la licencia MIT: consulte el archivo de LICENCIA para obtener más detalles.

Available Tools

2 tools
check_accessibilityC

Check web accessibility of a given URL using axe-core

ParametersJSON Schema
NameRequiredDescriptionDefault
urlYesURL to analyze
waitForSelectorNoOptional CSS selector to wait for before analysis
userAgentNoOptional user agent string to use for the request

TDQS

C2.9/5.0
Behavior2/5

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

With no annotations provided, the description carries the full burden of behavioral disclosure. It states what the tool does but doesn't describe how it behaves: it doesn't mention whether this is a read-only analysis, what the output format might be, potential rate limits, authentication requirements, or error conditions. For a tool that performs web analysis, this leaves significant behavioral gaps.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Conciseness5/5

Is the description appropriately sized, front-loaded, and free of redundancy?

The description is a single, efficient sentence that communicates the core purpose without any wasted words. It's appropriately sized for a tool with a clear, focused function and is front-loaded with the essential information.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Completeness2/5

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

Given that there are no annotations and no output schema, the description should provide more complete context for this accessibility checking tool. It doesn't explain what kind of results to expect, what accessibility standards are checked, whether the analysis is comprehensive or limited, or how the tool handles dynamic content. For a tool with 3 parameters and no structured output documentation, this is insufficient.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Parameters3/5

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

The input schema has 100% description coverage, so all parameters are documented in the schema itself. The description doesn't add any parameter-specific information beyond what's already in the schema descriptions. According to the scoring rules, when schema_description_coverage is high (>80%), the baseline is 3 even with no param info in the description.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Purpose4/5

Does the description clearly state what the tool does and how it differs from similar tools?

The description clearly states the tool's purpose with a specific verb ('Check') and resource ('web accessibility of a given URL'), and mentions the technology used ('axe-core'). However, it doesn't explicitly differentiate from its sibling tool 'simulate_colorblind', which appears to be a related but distinct accessibility function.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Usage Guidelines2/5

Does the description explain when to use this tool, when not to, or what alternatives exist?

The description provides no guidance on when to use this tool versus its sibling 'simulate_colorblind' or other alternatives. It doesn't mention prerequisites, typical use cases, or exclusions, leaving the agent with no contextual usage information beyond the basic purpose.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.

simulate_colorblindC

Simulate how a webpage looks for colorblind users

ParametersJSON Schema
NameRequiredDescriptionDefault
urlYesURL to capture
typeYesType of color blindness to simulate
outputPathNoOptional path to save the screenshot
userAgentNoOptional user agent string to use for the request

TDQS

C2.9/5.0
Behavior2/5

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

No annotations are provided, so the description carries the full burden of behavioral disclosure. It states the tool simulates colorblind views but doesn't describe how (e.g., generates a screenshot, modifies display, or returns data), what the output is (e.g., image file, visual report), or any behavioral traits like performance, rate limits, or side effects. This leaves significant gaps for an agent to understand the tool's operation.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Conciseness5/5

Is the description appropriately sized, front-loaded, and free of redundancy?

The description is a single, clear sentence: 'Simulate how a webpage looks for colorblind users.' It is front-loaded with the core purpose, has zero wasted words, and is appropriately sized for the tool's complexity. Every part of the sentence earns its place by conveying essential information efficiently.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Completeness2/5

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

Given the tool's moderate complexity (4 parameters, no output schema, no annotations), the description is incomplete. It lacks details on output (e.g., what is returned or saved), behavioral context (e.g., how simulation works, any limitations), and usage guidelines. While the schema covers parameters well, the description doesn't compensate for missing annotations or output schema, leaving the agent with insufficient context for effective use.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Parameters3/5

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

The input schema has 100% description coverage, clearly documenting all four parameters (url, type, outputPath, userAgent) with details like enum values for 'type.' The description doesn't add any parameter-specific information beyond what the schema provides, such as explaining the simulation process or output format. Given the high schema coverage, a baseline score of 3 is appropriate as the schema handles the heavy lifting.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Purpose4/5

Does the description clearly state what the tool does and how it differs from similar tools?

The description clearly states the tool's purpose: 'Simulate how a webpage looks for colorblind users.' It specifies the action (simulate) and resource (webpage appearance for colorblind users), making it easy to understand. However, it doesn't explicitly differentiate from its sibling tool 'check_accessibility,' which might also involve accessibility testing, though the focus here is specifically on colorblind simulation.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Usage Guidelines2/5

Does the description explain when to use this tool, when not to, or what alternatives exist?

The description provides no guidance on when to use this tool versus alternatives. It doesn't mention the sibling tool 'check_accessibility' or any other tools, nor does it specify prerequisites, contexts, or exclusions. Usage is implied from the purpose but lacks explicit direction.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.

Tool Schema Changelog

Recent tool additions, removals, and schema changes observed during successful MCP inspections.

  1. 2 tool updates
    • First observedcheck_accessibility
    • First observedsimulate_colorblind
Deploy Server
Give agents eyes on the webHosted screenshot MCP. Send a URL, get a pixel-perfect image. $0.002 per grab. Works with Claude, Cursor, and Codex.grabbit.liveAd

TDQS

B3.1/5.0

Scored across 2 tools

Disambiguation5/5

The two tools have clearly distinct purposes: one checks general web accessibility using axe-core, while the other specifically simulates colorblindness effects on a webpage. There is no overlap or ambiguity between them.

Naming Consistency5/5

Both tools follow a consistent verb_noun pattern (check_accessibility, simulate_colorblind) with clear, descriptive names. The naming style is uniform and predictable throughout the set.

Tool Count2/5

With only two tools, the server feels thin for a web accessibility domain. While the tools are useful, typical accessibility testing involves more operations like checking screen reader compatibility, keyboard navigation, or ARIA attributes, suggesting notable gaps in coverage.

Completeness2/5

The tool set is severely incomplete for web accessibility. It lacks core operations such as validating HTML structure, testing screen reader output, assessing keyboard accessibility, or generating accessibility reports, which are essential for comprehensive accessibility evaluation.

Maintenance

ActivityInactive
ResponsivenessNo issues

Resources

Related MCP Connectors

Related MCP Servers

Appeared in Searches