Resumen
Configuracion flujo de trabajo for configuracion up un single MCP gateway daemon ese fronts todos upstream MCP servidores. Cada AI assistant points at el daemon instead of spawning its own MCP fleet.
Desencadenador
One-time setup, plus rotation cada config cambio
Programación
One-time setup
Pasos del flujo de trabajo
1
Install MCP gateway
FastMCP, @modelcontextprotocol/server-gateway, o equivalent.
2
List upstream MCP servidores
Scavio MCP, Postgres MCP, GitHub MCP, internal MCPs.
3
Define gateway config
JSON listing cada upstream server's URL o command.
4
Ejecutar gateway as HTTP daemon
One procesar serving todos agents on un conocido port.
5
Point agents at el daemon
Claude Desktop, Cursor, opencode todos reference http://localhost:8765/mcp.
6
Verify consolidation
ps aux | grep mcp deberia mostrar 1, no 30+.
Implementacion en Python
Python
# Configuration-driven setup; no Python code needed.
# See gateway.json and per-agent mcp config below.Implementacion en JavaScript
JavaScript
// gateway.json
{
"upstreams": {
"scavio": { "url": "https://mcp.scavio.dev/mcp", "headers": { "x-api-key": "$SCAVIO_API_KEY" } },
"postgres": { "command": "npx", "args": ["@modelcontextprotocol/server-postgres", "$DATABASE_URL"] }
}
}
// Then in claude_desktop_config.json / .cursor/mcp.json / opencode config:
{
"mcpServers": {
"gateway": { "url": "http://localhost:8765/mcp" }
}
}Plataformas utilizadas
Búsqueda web con grafo de conocimiento, PAA y resúmenes de IA