🌐 Permita que los asistentes de IA naveguen y extraigan contenido de la web a través de una interfaz MCP sencilla.
El servidor MCP del navegador web proporciona a los modelos de IA la capacidad de explorar sitios web, extraer contenido y comprender páginas web mediante el Protocolo de Control de Mensajes (MCP). Permite la extracción inteligente de contenido con selectores CSS y una gestión robusta de errores.
🤝 Contribuir • 📝 Informar error
Related MCP server: mcp-web-tools
🎯 Extracción de contenido inteligente : seleccione exactamente lo que necesita con selectores CSS
⚡ Ultrarrápido : Construido con procesamiento asincrónico para un rendimiento óptimo
📊 Metadatos enriquecidos : captura títulos, enlaces y contenido estructurado
🛡️ Robusto y confiable : manejo de errores integrado y gestión de tiempos de espera
🌍 Multiplataforma : funciona en cualquier lugar donde se ejecute Python
Para instalar Web Browser Server para Claude Desktop automáticamente a través de Smithery :
npx -y @smithery/cli install web-browser-mcp-server --client claudeInstalar usando uv:
uv tool install web-browser-mcp-serverPara desarrollo:
# Clone and set up development environment
git clone https://github.com/blazickjp/web-browser-mcp-server.git
cd web-browser-mcp-server
# Create and activate virtual environment
uv venv
source .venv/bin/activate
# Install with test dependencies
uv pip install -e ".[test]"Agregue esta configuración a su archivo de configuración de cliente MCP:
{
"mcpServers": {
"web-browser-mcp-server": {
"command": "uv",
"args": [
"tool",
"run",
"web-browser-mcp-server"
],
"env": {
"REQUEST_TIMEOUT": "30"
}
}
}
}Para el desarrollo:
{
"mcpServers": {
"web-browser-mcp-server": {
"command": "uv",
"args": [
"--directory",
"path/to/cloned/web-browser-mcp-server",
"run",
"web-browser-mcp-server"
],
"env": {
"REQUEST_TIMEOUT": "30"
}
}
}
}El servidor proporciona una potente herramienta de navegación web:
Explorar y extraer contenido de páginas web con selectores CSS opcionales:
# Basic webpage fetch
result = await call_tool("browse_webpage", {
"url": "https://example.com"
})
# Target specific content with CSS selectors
result = await call_tool("browse_webpage", {
"url": "https://example.com",
"selectors": {
"headlines": "h1, h2",
"main_content": "article.content",
"navigation": "nav a"
}
})Configurar a través de variables de entorno:
Variable | Objetivo | Por defecto |
| Tiempo de espera de solicitud de página web en segundos | 30 |
Ejecute el conjunto de pruebas:
python -m pytestPublicado bajo la licencia MIT. Consulte el archivo de licencia para más detalles.
Hecho con ❤️ por el equipo de Pear Labs
Extract content from a webpage with optional CSS selectors for specific elements
| Name | Required | Description | Default |
|---|---|---|---|
| url | Yes | The URL of the webpage to browse | |
| selectors | No | Optional CSS selectors to extract specific content |
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 extracts content but fails to describe critical behaviors such as error handling (e.g., invalid URLs, network issues), performance traits (e.g., timeouts, rate limits), or output format. This leaves significant gaps in understanding how the tool operates beyond its basic function.
Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.
Is the description appropriately sized, front-loaded, and free of redundancy?
The description is a single, efficient sentence that front-loads the core purpose. It avoids redundancy and wastes no words, though it could be slightly more informative without sacrificing brevity. The structure is clear and direct, earning a high score for conciseness.
Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.
Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?
Given the tool's complexity (web scraping with optional selectors), lack of annotations, and no output schema, the description is incomplete. It doesn't explain what 'content' includes (e.g., text, links, structure), how selectors are applied, or potential limitations (e.g., JavaScript-rendered content). This leaves the agent with insufficient context for reliable use.
Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.
Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?
Schema description coverage is 100%, so the schema already documents both parameters ('url' and 'selectors') adequately. The description adds minimal value by mentioning 'optional CSS selectors for specific elements,' which aligns with the schema but doesn't provide additional syntax, examples, or constraints. This meets the baseline for high schema coverage.
Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.
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 ('Extract') and resource ('content from a webpage'), and mentions optional CSS selectors for refinement. It distinguishes the tool's core function effectively, though without sibling tools, differentiation isn't applicable. However, it lacks specificity about what 'content' entails (e.g., text, HTML, metadata), which prevents a perfect score.
Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.
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, prerequisites, or limitations. It mentions optional CSS selectors but doesn't explain when they are beneficial or necessary. With no sibling tools, context for usage is minimal, but the absence of any usage context results in a low score.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
browse_webpageScored across 1 tool
With only one tool, there is no possibility of ambiguity or overlap with other tools. The tool's purpose is clearly defined as extracting webpage content with optional element selection.
The single tool name follows a clear verb_noun pattern (browse_webpage). Since there are no other tools to compare against, consistency is inherently perfect.
A single tool for a web browser server is too minimal for the apparent scope. Basic web interactions like navigation, clicking, form filling, or handling multiple tabs are missing, making the server feel incomplete and limiting for agents.
The server is severely incomplete for a web browser domain. It only offers content extraction, lacking essential operations such as navigation, interaction with page elements, or session management, which are critical for typical web automation tasks.
Stealth web browser for agents: search, fetch, click, download and type in persistent MCP sessions.
Web search, browser automation, scraping, crawling and CAPTCHA solving for AI agents.
Web search, URL content extraction to Markdown, site mapping, and recursive web crawler.
Scrape, crawl and search the web for AI agents via MCP.