La búsqueda de empleo es repetitiva: busque las mismas palabras clave todos los días, consulte los mismos foros, investigue las mismas empresas. Con la búsqueda de MCP en Claude Desktop o Claude Code, puede automatizar la parte de descubrimiento y concentrarse en la aplicación. Este tutorial configura un flujo de trabajo diario de búsqueda de empleo.
Requisitos previos
- Claude Desktop o Claude Code con soporte MCP
- Una clave API de Scavio
Guia paso a paso
Paso 1: Configurar MCP para la búsqueda de empleo
Configure Scavio MCP para buscar ofertas de trabajo en Google y Reddit.
JSON
// .mcp.json or claude_desktop_config.json:
{
"mcpServers": {
"scavio": {
"url": "https://mcp.scavio.dev/mcp",
"headers": { "x-api-key": "your_scavio_api_key" }
}
}
}
// In Claude, you can now ask:
// 'Search Google for senior python developer remote jobs posted this week
// on Greenhouse and Lever. Also check Reddit for hiring threads.'Paso 2: Crear un mensaje de búsqueda diario
Cree un mensaje reutilizable para el descubrimiento diario de trabajos.
Python
# Save as a Claude Code custom command or paste daily:
JOB_SEARCH_PROMPT = """
Search for jobs matching my criteria:
- Role: Senior Python Developer
- Type: Remote
- Sources: Google (site:greenhouse.io, site:lever.co, site:jobs.ashbyhq.com)
- Also check: Reddit hiring threads
For each listing found:
1. Title and company
2. URL
3. Key requirements (if visible in snippet)
4. Whether it's new (not previously seen)
Format as a table.
"""Paso 3: Agregar investigación de empresa
Cuando encuentre listados interesantes, investigue la empresa mediante una búsqueda.
Bash
# After finding a listing, ask Claude:
# 'Research [Company Name] - search Google for their product, funding,
# and tech stack. Search Reddit for employee reviews and culture.'
# Claude will use:
# 1. google_search('[Company Name] funding tech stack')
# 2. reddit_search('[Company Name] working experience reviews')
# 3. Synthesize into a company brief
# This replaces manually checking Glassdoor, Crunchbase, etc.Paso 4: Seguimiento de solicitudes
Utilice Claude para mantener un rastreador de aplicaciones sencillo.
JSON
# Ask Claude Code to create/update a tracking file:
# 'Add [Company] [Role] to my job tracker. Status: applied.
# Mark the [Other Company] position as interviewed.'
# Claude Code can maintain a JSON or Markdown file like:
# applications.json:
# [
# {"company": "Acme", "role": "Senior Python Dev", "status": "applied", "date": "2026-05-06", "url": "..."},
# {"company": "TechCo", "role": "Backend Engineer", "status": "interviewed", "date": "2026-05-03", "url": "..."}
# ]Ejemplo en Python
Python
# This workflow runs entirely in Claude Desktop/Code via MCP.
# No custom Python needed - just configure .mcp.json and ask Claude.
# Example: 'Find remote Python jobs posted today on Greenhouse and Lever'Ejemplo en JavaScript
JavaScript
// MCP configuration only:
// { "mcpServers": { "scavio": { "url": "https://mcp.scavio.dev/mcp", "headers": {"x-api-key": "key"} } } }Salida esperada
JSON
An MCP-powered daily job search workflow that discovers listings from Google-indexed job boards, researches companies, and tracks applications.