runtime-mcp-server
Runtime Diagnostic MCP Server (runtime-mcp-server)
Un servidor local del Model Context Protocol (MCP) que ofrece a los agentes de codificación de IA (como OpenAI Codex CLI) diagnósticos del tiempo de ejecución en tiempo real, supervisión de procesos y análisis de logs para entornos de desarrollo locales.
🎯 Propósito
Los agentes de codificación de IA son excelentes inspeccionando código estático, pero quedan ciegos ante el estado de ejecución real. Este servidor cubre esa brecha dando a los agentes de IA herramientas programáticas y de solo lectura para observar procesos en ejecución, analizar logs de fallos, inspeccionar puertos activos y monitorizar los recursos del sistema.
Related MCP server: DevTwin MCP
🛠️ Herramientas principales (funciones de la v1)
tail_app_logsDescripción: Lee las últimas salidas de
stdout/stderro logs de fallos de los servidores de desarrollo activos en segundo plano (npm run dev,python,docker) y los sigue en tiempo real (tail).Objetivo: Entrega automáticamente las trazas de pila completas a Codex, sin necesidad de copiar y pegar manualmente.
inspect_portsDescripción: Escanea los puertos de red locales activos (p. ej., 3000, 8080) e identifica los identificadores de proceso (PIDs) que los ocupan.
Objetivo: Resuelve automáticamente los errores de
EADDRINUSE/ «El puerto ya está en uso» y puede enviar señales de forma segura a los procesos zombis.
get_process_metricsDescripción: Monitoriza el uso de CPU y memoria (RAM) de procesos concretos de la aplicación local.
Objetivo: Ayuda a Codex a detectar en tiempo real bucles infinitos, fugas de memoria y scripts en segundo plano descontrolados.
capture_network_errorsDescripción: Inspecciona las respuestas HTTP/gRPC locales fallidas, las cabeceras CORS y los códigos de estado.
Objetivo: Identifica las causas raíz de las conexiones de API locales rotas entre el frontend y el backend.
🏗️ Stack tecnológico
Lenguaje: TypeScript / Node.js
Protocolo: Model Context Protocol (MCP) mediante
@modelcontextprotocol/sdkTransporte: Entrada/Salida estándar (
stdio)APIs del sistema:
child_process,fs/promises,psutil/ utilidades de procesos del sistema
🚀 Primeros pasos
Instala las dependencias y compila el código fuente de TypeScript:
npm install
npm run buildArranca el servidor MCP:
npm startEl servidor inicial se conecta por stdio y queda listo para que se añadan herramientas de diagnóstico.
🚀 Integración con Codex CLI
Se configura en ~/.codex/config.json (o en los ajustes del cliente de Codex):
{
"mcpServers": {
"runtime-diagnostics": {
"command": "node",
"args": ["/Users/USERNAME/Documents/Codex/runtime-mcp-server/build/index.js"]
}
}
}📋 Hoja de ruta de desarrollo
Inicializar el proyecto TypeScript con el SDK
@modelcontextprotocol/sdkImplementar la herramienta
inspect_portsImplementar la herramienta
tail_app_logsImplementar la herramienta
get_process_metricsAñadir pruebas de integración con Codex CLI
Available Tools
1 toolinspect_portsA
Inspects active local network ports to determine which process ID (PID) and application are occupying a port.
| Name | Required | Description | Default |
|---|---|---|---|
| port | No | Optional local port to inspect. Must be an integer from 1 to 65535. |
TDQS
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
With no annotations, the description carries the full transparency burden. 'Inspects' signals a non-destructive read, and the result (PID/application) is stated. However, it does not disclose what happens when port is omitted, what output shape to expect, or whether elevated privileges are required.
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 focused sentence that states the action, target resource, and purpose without filler or repetition. Every phrase contributes useful information.
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?
The tool is simple, but with no output schema and no annotations, the description should clarify omitted-port behavior and return contents more precisely. Mentioning PID and application is likely enough for basic use, but some interpretation remains.
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?
The schema already fully describes the port parameter, including optionality and range. The description adds no new parameter semantics beyond the association between port being and PID/application resolution, so the baseline score of 3 is appropriate.
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 uses a specific verb ('Inspects'), names the resource ('active local network ports'), and states the outcome (identify PID and application). It clearly distinguishes itself from any generic or unknown tool, and no sibling tools are provided to confuse selection.
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 gives clear context: use this tool when you need to map a local port to the process and app using it. It does not explicitly state when not to use it or name alternatives, but the absence of sibling tools makes that less necessary.
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 tool update
v1.0.0- First observed
inspect_ports
TDQS
Scored across 1 tool
With only one tool available, there is no possibility of confusing it with other tools. The tool's purpose is clearly defined and stands alone.
The single tool name follows a clear verb_noun pattern. Since there is only one tool, there are no inconsistencies or competing naming conventions to evaluate.
The server is named 'runtime-mcp-server,' suggesting broad runtime management capabilities, but it exposes only a single port inspection tool. This feels severely undersized for the stated scope.
The tool surface is limited to inspecting ports, with no support for managing processes, terminating processes, or interacting with runtime state in any other meaningful way. The server's purpose appears very narrow, leaving significant gaps for any real runtime management workflow.
Maintenance
Related MCP Connectors
Deploy local apps, add managed Postgres, and inspect logs from your AI coding agent.
Your team's shipping standards, org map and delivery metrics, inside your coding agent.
- SuperlogOAuthsh.superlog
Open-source agent that observes and fixes your application. Query logs, traces, metrics, incidents.
AI agent observability for production traces, natural-language insights, and improvement loops.
Related MCP Servers
- AlicenseNot gradedqualityDmaintenanceEnables LLM clients to inspect local dev environments—Docker container health, pnpm workspace integrity, and stuck process detection—without manual terminal copy-pasting.MIT
- AlicenseAqualityBmaintenanceProvides AI coding agents with structured, evidence-based diagnostics about the local development environment, detecting tech stack, runtime mismatches, dependency state, services, ports, and Git status without exposing secrets or using network calls.10Apache 2.0
- AlicenseNot gradedqualityCmaintenanceExposes a live application's runtime state to coding agents, including logs, DB schema, config, and feature flags, with read-only SQL and secret redaction.MIT
- AlicenseNot gradedqualityBmaintenanceEnables coding agents to access live runtime observability data such as logs, deploys, and health metrics for evidence-based incident triage.1MIT