iron-mcp-framework: Skill for Codex

.agents/skills/create-mcp-tool/SKILL.md

create-mcp-tool is a skill for Codex from julioalberto64/iron-mcp-framework. It costs 69 tokens per session (1,524 once invoked), scanned A, original, MIT.

A development workflow for adding MCP tools to the IronMCP framework. MCP, or Model Context Protocol, is a standard way for an AI client to call external tools; this workflow covers schemas, permissions, cancellation, registration, tests, and examples.

In plain words
What is it for?
Use it to create or extend a tool backed by a local operation or REST endpoint in IronMCP. It helps place the code correctly, connect it to the registry, define inputs and permissions, and validate the result.
Why use it?
It replaces guesswork with the repository's required implementation and validation steps. It also tells the agent to inspect the current code and ask for missing integration details instead of inventing them.

Skill for Codex

Written for Codex: agents/openai.yaml present. Also seen: installed under .agents/ (shared by several agents).

This is julioalberto64/iron-mcp-framework's own configuration. It tells Codex how to work on iron-mcp-framework itself, so it is not a mod to install elsewhere. Copy it as a starting point and replace the rules that are about this project. Everything iron-mcp-framework configures →

Reuse

Borrowing it

Nothing to install: this file belongs to julioalberto64/iron-mcp-framework. Take a copy, put it at the same path in your own repository, and replace the rules that are about this project with yours.

Copy the file
curl -O https://raw.githubusercontent.com/julioalberto64/iron-mcp-framework/main/.agents/skills/create-mcp-tool/SKILL.md
Clone the repo
git clone --depth 1 https://github.com/julioalberto64/iron-mcp-framework

Made for: Codex.

Wrote this? Show the measurements

A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.

agentmods badge for create-mcp-tool

README.md
[![agentmods](https://agentmods.dev/badge/skills/julioalberto64/iron-mcp-framework/create-mcp-tool/github.svg)](https://agentmods.dev/skills/julioalberto64/iron-mcp-framework/create-mcp-tool)
Your own site
<a href="https://agentmods.dev/skills/julioalberto64/iron-mcp-framework/create-mcp-tool"><img src="https://agentmods.dev/badge/skills/julioalberto64/iron-mcp-framework/create-mcp-tool/github.svg" alt="Measured on agentmods" height="20"></a>

Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.

agentmods 80×15 button for create-mcp-tool

Your own site · 80×15
<a href="https://agentmods.dev/skills/julioalberto64/iron-mcp-framework/create-mcp-tool"><img src="https://agentmods.dev/badge/skills/julioalberto64/iron-mcp-framework/create-mcp-tool.svg" alt="Reviewed on agentmods" width="80" height="20"></a>
Per session 69 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 1,524 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. A grade says what 26 rules found in the file — not that it is safe.
Origin original No closer match found in the catalogue.
Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

ModelPer sessionOnce invoked
Fable 5.1 $0.00069 $0.01524
Opus 5.5 $0.00028 $0.00610
Sonnet 5.5 $0.00014 $0.00305
Haiku 4.5 $0.00007 $0.00152

Measured 17d ago against content hash 5d535ef1dc6d, method: parsed. Prices are Anthropic first-party input rates as of 2026-10-07, from the pricing page.

Security

Grade A, and why

create-mcp-tool scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 17d ago.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Nothing flagged

None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.

.agents/skills/create-mcp-tool/SKILL.md · 55 lines

How it starts

The opening of the file, as written. The whole thing — 55 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Crear una tool en IronMCP

Implementa la tool solicitada y deja el cambio validado. Las rutas src/, test/, examples/ y docs/ de esta guía son relativas a la raíz del repositorio. Da prioridad al código vigente si esta guía y el contrato difieren.

Ubicar el contrato

Lee src/framework/types.ts, src/framework/create-mcp-server.ts, src/tools/index.ts y una tool similar antes de editar. Para una integración REST, consulta también src/framework/rest/rest-client.ts. Usa references/example.md cuando necesites un ejemplo completo de tool, prueba y petición HTTP; adapta sus campos al endpoint real.

A partir del pedido y la documentación disponible, identifica nombre, propósito, entradas, operación, respuesta esperada y permisos. Si falta un dato que impida definir la integración, pide ese dato; no inventes endpoints ni campos del proveedor. Si la información alcanza, implementa sin detenerte en una propuesta.

Implementar e integrar

  • Crea src/tools/<nombre-kebab-case>.tool.ts con un nombre MCP único en snake_case y defineTool de ../framework/types.js. Usa imports .js para módulos locales e import * as z from "zod/v4".
  • Define name, title, description, inputSchema, outputSchema, timeoutMs, idempotent, auth, audit y handler. outputSchema es opcional en TypeScript, pero es la convención para nuevas tools con salida estructurada. Los schemas de registro son shapes, no z.object(...); puedes compartir .shape con los objetos usados para validar.
  • Declara timeoutMs como un literal entre 1 y 30000; el validador actual inspecciona texto. Describe con precisión si la operación consulta o cambia estado y define idempotent según su comportamiento real, no solo por el verbo HTTP.
  • Valida dentro de handler(input, context) con Zod antes de acceder a la API. Restringe formato y longitud según el dominio; codifica cada segmento dinámico con encodeURIComponent y construye las queries con URLSearchParams.
  • Para REST usa context.restClient.get o .post. Este cliente está asociado al AbortSignal de la ejecución. No captures un cliente global ni crees otro cliente dentro de la tool. Si necesitas otro método HTTP que el cliente todavía no ofrece, extiéndelo con el cambio mínimo y sus pruebas; no simules el método ni eludas el cliente.
  • Fija la ruta según el propósito de la tool. No aceptes URLs completas ni rutas arbitrarias del usuario. El cliente admite únicamente rutas del origen configurado, bloquea redirecciones y limita las respuestas a 2 MiB.
  • Valida las respuestas externas relevantes: los genéricos TypeScript no validan JSON. Devuelve un objeto acorde a outputSchema, con los campos necesarios para el usuario. No incluyas raw, headers, tokens o datos personales ajenos al propósito por copiar otras tools.
  • El framework transforma el objeto devuelto en un resultado MCP; no construyas manualmente content ni isError dentro del handler. Propaga los fallos; usa los errores existentes de src/framework/errors.ts cuando aporten contexto seguro.
  • Registra la tool en el array devuelto por createTools() en src/tools/index.ts, conservando las existentes. Una factory, si resulta útil, no debe capturar el RestClient: el handler lo obtiene del contexto.

Read the full file on GitHub · 55 lines

Files

What ships with it

2 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.

Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

  1. 17d ago First seen · 55 lines · 69 tokens per session scan A 5d535ef1dc6d

Subscribe to this mod's changes

create-mcp-tool is a skill published in the GitHub repository julioalberto64/iron-mcp-framework (1 stars, last pushed 18d ago), licensed MIT. It adds 69 tokens to every session and 1,524 once invoked, about $0.0003 per session on Opus 5.5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-10-20.

Related

create-mcp-tool compared with skills from other repositories

create-mcp-tool next to 6 skills from other repositories
ModTypeStarsLast pushLicenceScanTokensWorks with
create-mcp-tooljulioalberto64/iron-mcp-frameworkSkill118d agoMITA69Codex
mem0-oss-to-platformmem0ai/mem0Skill68k2d agoApache-2.0A217Claude Code, Codex
azure-mgmt-botservice-dotnetmicrosoft/skillsSkill3.2kyesterdayMITA78Claude Code
fastapi-router-pymicrosoft/skillsSkill3.2kyesterdayMITA46Claude Code
dev-runtimelangwatch/langwatchSkill5.0kyesterdayApache-2.0A172Claude Code
api-designericrisco/rsc-harnessSkill5032d agoMITA105Claude Code, Codex
migrate-segw-to-raparc-mcp/arc-1Skill252yesterdayMITA106Claude Code