Skip to content

el-buen-agente-mcp server

io.github.apasztetnik/el-buen-agente-mcpapasztetnikstreamable-httpv2.8.1Scanned

La guía 'El Buen Agente' como 18 tools para evaluar, mejorar y construir agentes LLM. En español.

18 tools: 17 change things and 1 only read.

Tools: 0 delete, 0 irreversible, 17 write, 1 read.
  • Reads private data · 0
  • Reads text others write · 0
  • Can send data out · 0
  • No findings

Tools listed without sign-in. Remote posture

With Anlyon

17 tools here change things. Govern them.

  • Anlyon lists the tools and adopts none.
  • Each tool you adopt gets approvals and receipts.
  • Your agent never holds the key.
Govern this server →

A tool your agent calls directly is not routed through Anlyon and is not governed by it. Anlyon Radar only lists and labels, and these labels are the same either way.

What the el-buen-agente-mcp server can do

The el-buen-agente-mcp server is published by the GitHub account apasztetnik. It has 18 tools. It can create or change data (17 tools: recomendar_flujo, evaluar_necesidad, revisar_rol_y_frontera, revisar_outputs and 13 more), but cannot delete data. 1 tool only reads.

None of its tools reads private data, brings in text other people write, or sends data out.

None of the fixed checks matched anything in its tool definitions. That is not a safety guarantee.

Its remote endpoint lists its tools without sign-in.

Last scanned Nov 16, 2026, version 2.8.1.

Tools

Grouped by what each tool does, most consequential first. Open one for its reasons and full description; descriptions are the publisher's own words.

18 tools

Changes things 17Claude Code: ask
  • aplicar_challengerAplica el patrón challenger/red-team a la definición o a una decisión del agente: contraargumentos basados en datos, autocrítica y gate de calidad.No findings

    Why it is in this group

    • Nothing in its name or description says what it does, so it is counted as a write.
    Description, as the server wrote it
    Aplica el patrón challenger/red-team a la definición o a una decisión del agente: contraargumentos basados en datos, autocrítica y gate de calidad. Útil como segunda pasada adversarial. Recibe la definición del agente y devuelve un brief de evaluación estructurado.
  • auditar_contextoAudita el contexto del agente: separación en 3 capas (identidad/dominio/referencia), relevancia y caducidad de datos, estrategia de integración (directo/snapshot/RAG/estático), gob…No findings

    Why it is in this group

    • Nothing in its name or description says what it does, so it is counted as a write.
    Description, as the server wrote it
    Audita el contexto del agente: separación en 3 capas (identidad/dominio/referencia), relevancia y caducidad de datos, estrategia de integración (directo/snapshot/RAG/estático), gobernanza, least-privilege, trazabilidad y defensa anti prompt-injection. Recibe la definición del agente y devuelve un brief de evaluación estructurado.
  • challenger_decisionGenera el brief para cuestionar una decisión concreta del agente: 3 razones basadas en datos para NO hacerla.No findings

    Why it is in this group

    • Nothing in its name or description says what it does, so it is counted as a write.
    Description, as the server wrote it
    Genera el brief para cuestionar una decisión concreta del agente: 3 razones basadas en datos para NO hacerla. No bloquea, informa al humano.
  • checklist_nacimientoCorre el checklist completo de 19 puntos contra la definición del agente.No findings

    Why it is in this group

    • Nothing in its name or description says what it does, so it is counted as a write.
    Description, as the server wrote it
    Corre el checklist completo de 19 puntos contra la definición del agente. Es el gate final antes de mergear: el agente debe NACER cumpliéndolo, no corregirse después. Devuelve veredicto punto por punto.
  • construir_agenteEl paso de CIERRE del flujo: toma la definición iterada (tras pasar por las revisiones) y produce el artefacto final listo para usar: identity layer, tools con least-privilege, lím…No findings

    Why it is in this group

    • Nothing in its name or description says what it does, so it is counted as a write.
    Description, as the server wrote it
    El paso de CIERRE del flujo: toma la definición iterada (tras pasar por las revisiones) y produce el artefacto final listo para usar: identity layer, tools con least-privilege, límites duros, schema de output, gates, contexto, evaluación. Llamala cuando checklist_nacimiento dé APTO.
  • disenar_evaluacionEvalúa el plan de evaluación del agente (o ayuda a crearlo): 3 dimensiones (capacidades/trayectoria/resultado), métricas (tasa de éxito, consistencia, coste por tarea, adopción), g…No findings

    Why it is in this group

    • Nothing in its name or description says what it does, so it is counted as a write.
    Description, as the server wrote it
    Evalúa el plan de evaluación del agente (o ayuda a crearlo): 3 dimensiones (capacidades/trayectoria/resultado), métricas (tasa de éxito, consistencia, coste por tarea, adopción), golden set de 20-50 tareas, monitoreo de drift y self-consistency para alto stake.
  • evaluar_autonomiaDetermina el nivel de autonomía adecuado (copiloto / ejecutor supervisado / autónomo con guardrails) y verifica los mecanismos de reducción de riesgo: sandbox/shadow mode, límites …No findings

    Why it is in this group

    • Nothing in its name or description says what it does, so it is counted as a write.
    Description, as the server wrote it
    Determina el nivel de autonomía adecuado (copiloto / ejecutor supervisado / autónomo con guardrails) y verifica los mecanismos de reducción de riesgo: sandbox/shadow mode, límites duros, frenos progresivos, override humano. Recibe la definición del agente y devuelve un brief de evaluación estructurado.
  • evaluar_necesidadEvalúa si el problema justifica un agente o se resuelve con menos (prompt, workflow, skill).No findings

    Why it is in this group

    • Nothing in its name or description says what it does, so it is counted as a write.
    Description, as the server wrote it
    Evalúa si el problema justifica un agente o se resuelve con menos (prompt, workflow, skill). Detecta antipatrones: agente genérico, sobre-orquestación, agente sin contexto, autonomía total día 1. Usala ANTES de construir, o para cuestionar un agente existente.
  • evaluar_sistemaEvalúa cómo el agente encaja en el sistema mayor: catálogo de skills, reutilización vs especialización, orquestación ligera vs orquestador, qué automatizar vs supervisar, y memoria…No findings

    Why it is in this group

    • Nothing in its name or description says what it does, so it is counted as a write.
    Description, as the server wrote it
    Evalúa cómo el agente encaja en el sistema mayor: catálogo de skills, reutilización vs especialización, orquestación ligera vs orquestador, qué automatizar vs supervisar, y memorias separadas por agente.
  • generar_contratoGenera el contrato formal del agente (patrón contractor de §8) a partir de los campos provistos.No findings

    Why it is in this group

    • Nothing in its name or description says what it does, so it is counted as a write.
    Description, as the server wrote it
    Genera el contrato formal del agente (patrón contractor de §8) a partir de los campos provistos. Los campos faltantes quedan marcados como [PENDIENTE] para completar.
  • plan_de_inicioGenera el plan de arranque correcto para un agente nuevo: un agente, un problema concreto, un humano revisando.No findings

    Why it is in this group

    • Nothing in its name or description says what it does, so it is counted as a write.
    Description, as the server wrote it
    Genera el plan de arranque correcto para un agente nuevo: un agente, un problema concreto, un humano revisando. Verifica que la tarea elegida sea recurrente, costosa en tiempo y con datos accesibles.
  • plan_exposicion_mcpEvalúa qué partes del agente conviene exponer a la economía de agentes y cómo: qué modelar como tool (capacidad accionable), resource (doc/dato legible) o prompt (plantilla), y qué…No findings

    Why it is in this group

    • Nothing in its name or description says what it does, so it is counted as a write.
    Description, as the server wrote it
    Evalúa qué partes del agente conviene exponer a la economía de agentes y cómo: qué modelar como tool (capacidad accionable), resource (doc/dato legible) o prompt (plantilla), y qué merece UI vs API/MCP.
  • recomendar_flujoDevuelve el flujo recomendado de tools según la situación: agente nuevo (diseño desde cero) o agente existente (mejora).No findings

    Why it is in this group

    • Nothing in its name or description says what it does, so it is counted as a write.
    Description, as the server wrote it
    Devuelve el flujo recomendado de tools según la situación: agente nuevo (diseño desde cero) o agente existente (mejora). Llamala PRIMERO si no sabés por dónde empezar.
  • revisar_frontera_ejecucionVerifica que la línea entre lo que el agente recomienda y lo que ejecuta esté definida en código (status + gates), no descubierta en producción: qué ejecuta directo, qué queda pend…No findings

    Why it is in this group

    • Nothing in its name or description says what it does, so it is counted as a write.
    Description, as the server wrote it
    Verifica que la línea entre lo que el agente recomienda y lo que ejecuta esté definida en código (status + gates), no descubierta en producción: qué ejecuta directo, qué queda pendiente de aprobación, qué se rechaza. Recibe la definición del agente y devuelve un brief de evaluación estructurado.
  • revisar_outputsVerifica que los outputs sean accionables: schema estricto (JSON Schema/Pydantic), resumen legible para humanos separado del razonamiento, y exposición de qué gates pasó/falló.No findings

    Why it is in this group

    • Nothing in its name or description says what it does, so it is counted as a write.
    Description, as the server wrote it
    Verifica que los outputs sean accionables: schema estricto (JSON Schema/Pydantic), resumen legible para humanos separado del razonamiento, y exposición de qué gates pasó/falló. Recibe la definición del agente y devuelve un brief de evaluación estructurado.
  • revisar_rol_y_fronteraVerifica que el agente tenga rol claro, dominio acotado y frontera explícita de qué NO es su responsabilidad, escrita en el identity layer y enforced en código donde se pueda.No findings

    Why it is in this group

    • Nothing in its name or description says what it does, so it is counted as a write.
    Description, as the server wrote it
    Verifica que el agente tenga rol claro, dominio acotado y frontera explícita de qué NO es su responsabilidad, escrita en el identity layer y enforced en código donde se pueda. Recibe la definición del agente y devuelve un brief de evaluación estructurado.
  • validar_veredictoCierra el ciclo de checklist_nacimiento con un contrato a nivel protocolo.No findings

    Why it is in this group

    • Nothing in its name or description says what it does, so it is counted as a write.
    Description, as the server wrote it
    Cierra el ciclo de checklist_nacimiento con un contrato a nivel protocolo. Pasale los 19 puntos con su estado (ok|parcial|falta) y devuelve structuredContent validado: conteos y veredicto normalizado (apto solo si faltas === 0). Pensada para consumo programático / gates de CI: no depende de parsear texto.
Only reads 1Show 1 read-only tool
  • get_el_buen_agenteDevuelve la guía completa 'El Buen Agente (v2)' en Markdown.No findings

    Why it is in this group

    • The name "get_el_buen_agente" says "get": a word for reading.
    Description, as the server wrote it
    Devuelve la guía completa 'El Buen Agente (v2)' en Markdown. Usala para contexto general; para operar sobre un agente concreto usá las tools evaluar_*/revisar_*/auditar_*.

This is not a safety rating, and no finding is not a safety guarantee. The groups and findings come from fixed checks of the tool definitions the server returned on the date shown, not from its code or what it does when called. How the checks work.

Exfil surface

Which tools read private data, which read text someone else can write, and which can send data out. One of each is a path a prompt injection can use.

No private-data, untrusted-input, outbound path in this server alone: no tool is tagged private-data or untrusted-input or outbound.

Reads private data0
Reads files, mail, messages, databases, secrets, calendars or repositories.
Reads text others write0
Brings in text someone else can write: web pages, inbound email, messages, issues and comments.
Can send data out0
Can carry data to people or systems: it sends, posts or publishes, or takes a recipient, channel or URL.

Other servers in the same agent add their own tools to these paths. Audit your whole setup to see them together.

Remote posture

What anyone on the internet sees at the remote endpoint: an anonymous initialize and tools/list, and the public OAuth metadata.

Tools listed without sign-in

Looked at Nov 20, 2026, 19:02 UTC, unchanged since Oct 22, 2026. No credentials were sent and no tool was called. What is looked at.

Reachable
yes
HTTP status of an anonymous initialize
200
tools/list served without credentials
yes
Tools that change things, listed without credentials
17
Protocol version
2025-06-18
WWW-Authenticate on 401
not observed
resource_metadata in WWW-Authenticate
not observed
RFC 9728 protected resource metadata
not observed
Authorization servers
not observed
Legacy HTTP+SSE transport
no
Dynamic client registration
not observed
Client ID metadata documents
not observed
  • WarnThe whole server

    The endpoint served tools/list without credentials, and the list has 17 tools that change things (17 write). Whether a call also works without credentials was not tested: tools are never called.

    Matched text

    https://el-buen-agente-mcp-production.up.railway.app/mcp: tools/list returned 18 tools to an anonymous client, including recomendar_flujo (write), evaluar_necesidad (write), revisar_rol_y_frontera (write), revisar_outputs (write), evaluar_autonomia (write) and 12 more

    What this check looks for

Findings

Matches of fixed checks, with the text they matched.

No check matched anything in these tool definitions. That is not a safety guarantee.

History

Changes to its tool definitions and to its remote endpoint, newest first.

No change seen yet. Its tools are rescanned every day, and a remote endpoint is looked at once a week.

Install

Pinned to version 2.8.1. Values in angle brackets are yours to fill in.

claude mcp add --transport http el-buen-agente https://el-buen-agente-mcp-production.up.railway.app/mcp
claude mcp add --transport http el-buen-agente https://el-buen-agente-mcp-production.up.railway.app/mcp

Run in a terminal.

Claude Code permissions

Allow the reads, ask before a change or a way out, deny the deletes. Paste into .claude/settings.json.

.claude/settings.json
{
  "permissions": {
    "allow": [
      "mcp__el-buen-agente__get_el_buen_agente"
    ],
    "ask": [
      "mcp__el-buen-agente__recomendar_flujo",
      "mcp__el-buen-agente__evaluar_necesidad",
      "mcp__el-buen-agente__revisar_rol_y_frontera",
      "mcp__el-buen-agente__revisar_outputs",
      "mcp__el-buen-agente__evaluar_autonomia",
      "mcp__el-buen-agente__revisar_frontera_ejecucion",
      "mcp__el-buen-agente__aplicar_challenger",
      "mcp__el-buen-agente__auditar_contexto",
      "mcp__el-buen-agente__challenger_decision",
      "mcp__el-buen-agente__disenar_evaluacion",
      "mcp__el-buen-agente__generar_contrato",
      "mcp__el-buen-agente__evaluar_sistema",
      "mcp__el-buen-agente__plan_exposicion_mcp",
      "mcp__el-buen-agente__checklist_nacimiento",
      "mcp__el-buen-agente__validar_veredicto",
      "mcp__el-buen-agente__construir_agente",
      "mcp__el-buen-agente__plan_de_inicio"
    ],
    "deny": []
  }
}

Tools that only read are allowed, tools that change things ask first, tools that delete or cancel are denied. A read that can send data out (it takes a URL, say) asks first too. It assumes the server is installed as el-buen-agente. Review it before you use it.

Pin

Keep the decision: fail CI when the tools change, follow every change, or show a badge.

Pin in CI

Fail a build when the tools you reviewed are not the tools the server serves today.

curl -s "https://radar.anlyon.com/api/v1/servers/io.github.apasztetnik/el-buen-agente-mcp/verify?hash=de984b5dabc02fc33cac297e64d91e0601c251d7eaa31b8ad0d848c39cd5d32d" | jq -e '.changed == false'

Pinned hash de984b5dabc0, from Nov 16, 2026, 10:17 UTC.

Watch

Get told when its tools change: a new tool, a new description, a new class.

https://radar.anlyon.com/servers/io.github.apasztetnik/el-buen-agente-mcp/changes.xml

An Atom feed. Works in any feed reader or Slack's RSS app.

Badge

Anlyon Radar badge for io.github.apasztetnik/el-buen-agente-mcp
[![Anlyon Radar](https://radar.anlyon.com/badge/io.github.apasztetnik/el-buen-agente-mcp.svg)](https://radar.anlyon.com/servers/io.github.apasztetnik/el-buen-agente-mcp)

Source: MCP Registry · repository. Synced Oct 21, 2026, 18:23 UTC.

el-buen-agente-mcp server: questions and answers

Answered from the scan above, not from the publisher's description.

Is the el-buen-agente-mcp server safe?

No scan can promise that. What its definitions show: 18 tools, 17 of which can change, send or delete something, and 0 findings from fixed checks. Allow the reads, ask before each change, and deny the deletes; the permissions block on this page does that.

Can the el-buen-agente-mcp server delete data?

No tool is classed as destructive: going by their definitions, none of them deletes, cancels or revokes anything.

Can the el-buen-agente-mcp server do things that cannot be undone?

No tool is classed as irreversible: none of them sends, pays, publishes or merges, going by their definitions.

What tools does the el-buen-agente-mcp server have?

18 tools: 17 write, 1 read. recomendar_flujo, evaluar_necesidad, revisar_rol_y_frontera, revisar_outputs, evaluar_autonomia, revisar_frontera_ejecucion, aplicar_challenger, auditar_contexto and 10 more.

Does the el-buen-agente-mcp server need sign-in or OAuth?

Its remote endpoint listed its tools to an anonymous client, without sign-in.

Does the el-buen-agente-mcp server send data to outside services?

None of its tools sends, posts or publishes, or takes a recipient, channel or URL.

Has the el-buen-agente-mcp server changed its tools?

Not since Anlyon Radar started tracking it. Last scanned Nov 16, 2026.