Spec Driven Development
de formulahendry
MCP Server that facilitates spec-driven development workflows, not just Vibe Coding.
Se ejecuta en tu computadora★ 440MITActualizado hace 14 meses
Instalar
MX Stage
Resumen
Descripción del creador (original en inglés)
Local AI workbench for IBM Maximo data: every change is staged, diffed and approved before commit.
Los tres pilares de PopNx MCP
Estos tres nunca se combinan en una sola puntuación: las opiniones, las mediciones y la popularidad son cosas distintas.
- Opinión de usuariosiAún no hay reseñasLas reseñas se abrirán junto con las cuentas en una etapa posterior
- Nuestras revisiones básicasi1 de 1 superadasRevisado con información pública; el código aún no se ha escaneado
- UsoiAún sin medir★ –
Qué puedes hacer con él
Todavía no escribimos un resumen sencillo para este. Abajo están los datos que leímos del registro.
- Conexión: Se ejecuta en tu computadora
- Transporte: stdio
- Categoría: Otros
Instalación
Elige la app de IA que usas y copia la configuración.
Antes de instalar
- Revisiones básicas: 1 de 1 superadas
- Se ejecuta en tu computadora con los permisos de tu usuario (archivos, red, apps).
- El registro no indica ninguna clave que completar.
Necesitas
- npmNode.js instalado (para el comando npx)
¿Qué app usas?
Elige tu app arriba. La recordamos, así que la próxima vez el botón dirá “Instalar en …”.
Descargar el archivo .mcpb del creador
El enlace lleva al archivo del propio creador; no lo alojamos ni lo reempaquetamos. Abre el archivo y Claude Desktop lo instalará.
Ejecuta esto en tu terminal
claude mcp add mxstage -- npx -y @mxstage/mxstageO copia la configuración tú
Pégalo en .cursor/mcp.json de tu proyecto, o en ~/.cursor/mcp.json
{
"mcpServers": {
"mxstage": {
"command": "npx",
"args": [
"-y",
"@mxstage/mxstage"
]
}
}
}Tu navegador te pide permiso para abrir la app; luego la app muestra la configuración y te pide confirmar. Los valores de las claves quedan vacíos para que los completes tú.
O copia la configuración tú
Pégalo en .vscode/mcp.json de tu proyecto
{
"servers": {
"mxstage": {
"type": "stdio",
"command": "npx",
"args": [
"-y",
"@mxstage/mxstage"
]
}
}
}Tu navegador te pide permiso para abrir la app; luego la app muestra la configuración y te pide confirmar. Los valores de las claves quedan vacíos para que los completes tú.
Este se ejecuta en tu computadora, y ChatGPT solo puede conectarse a servidores en internet. Elige otra app.
Tu elección se guarda en una cookie en este dispositivo; no necesitas cuenta.
Esta configuración se genera a partir de la ficha del creador en el registro. Si el README dice otra cosa, sigue el README.
Qué puede hacer
Se ejecuta en tu computadora; la lista de herramientas aparece en tu app de IA después de instalarlo.
Revisiones básicasi
Revisamos información que cualquiera puede ver; no hemos leído el código. Las revisiones que no aplican no se cuentan. · Revisado el 6 nov 2026
- ✓ Superada
El creador coincide con el repo
El nombre en el registro se verificó con la cuenta de GitHub mxstage, dueña del repo
- … Aún sin revisar
Sigue en mantenimiento
Todavía no hemos leído este repo en GitHub
- … Aún sin revisar
Tiene licencia
Todavía no hemos leído este repo en GitHub
- – No aplica
El endpoint responde
Este se ejecuta en tu computadora; no hay endpoint en internet
- – No aplica
Inicio de sesión con OAuth
Se ejecuta en tu computadora y usa ajustes locales en su lugar
- … Aún sin revisar
Sin instrucciones ocultas en las descripciones de herramientas
Se ejecuta en tu computadora; solo podemos leer listas de herramientas de endpoints en internet
Comparado con otros en Otrosi
- Más estrellas que el 0% de los otros 32384 servidores de esta categoría
- Publicado en el registro más recientemente que el 69% de los otros 32384 servidores de esta categoría
- Revisiones básicas iguales o mejores que el 100% de los otros 32384 servidores de esta categoría
Comparado con todos los servidores activos de esta categoría en el registro. Las estrellas solo cuentan cuando quien lo publica es dueño del repo.
Preguntas frecuentes
¿Es seguro MX Stage?
Superó 1 de 1 revisiones básicas aplicables. No hemos leído el código. Antes de agregar una clave o un token, revisa qué acceso pide el servidor y usa una clave que tenga solo los permisos que necesitas.
¿Qué claves o códigos necesita?
El registro no indica ninguna clave obligatoria, aunque algunos servidores te piden iniciar sesión la primera vez que los usas.
¿Se ejecuta en mi computadora o en la nube?
Se ejecuta en tu computadora. Tu app de IA lo inicia cuando hace falta, y los datos no pasan por los servidores del creador (a menos que el propio servidor llame a un servicio en línea).
¿Sigue en mantenimiento?
El repo de GitHub cambió por última vez –, y la última versión (0.2.9) llegó al registro hace 17 días.
¿Qué apps pueden usarlo?
Las apps compatibles con MCP, como Claude Desktop, Claude Code, Cursor y VS Code. También necesitas MCPB, Node.js en tu computadora.
¿Qué puedo usar en su lugar?
En la misma categoría: Spec Driven Development, Remote Desktop Commander, 오늘냥이 (Oneul Nyangi)
Alternativas
Servidores de la misma categoría hechos por otros creadores, ordenados por estrellas.
Cómo leer una tarjeta:★ Estrellas de GitHubiTres pilares: reseñas · revisiones · usoiActualizado = fecha en que se publicó la última versión en el registroi
Remote Desktop Commander
de desktop-commander
Remote Desktop Commander: use files, terminal and processes on your own computers from any AI app.
Se ejecuta en la nube★ 211Actualizado ayer
Instalar
오늘냥이 (Oneul Nyangi)
de nyangi.today
Real cat photos, diaries, profiles, age and food reference calculators, and breeds.
Se ejecuta en la nubeActualizado hoy
Instalar
README del creador
Abrir el README
Copiado tal cual del repo del creador. PopNx MCP no lo ha revisado ni lo respalda. Se quitaron HTML, imágenes y scripts. Los enlaces llevan fuera de este sitio.
MX Stage
English | 日本語
MX Stage is a local-only AI workbench for correcting IBM Maximo data with your own AI assistant (Claude Desktop, ChatGPT, Antigravity, IBM Bob): every change is staged, diffed and approved by a human before it is written to Maximo.
Website: https://mxstage.tsunagi.app · Contact: mxstage@tsunagi.app
- Your AI works on a staging sheet, not on Maximo. The AI assistant loads Maximo data into a sheet in the *work screen* (a browser tab), analyses it and proposes changes with reasons. You see every changed cell. The sheets live only in the work screen tab and the bridge's memory; loaded tables are never written to disk.
- Only a person can commit. The AI's tools cannot write to Maximo. Changes are written only when you press Commit to Maximo in the work screen. MX Stage writes one record first and lets you check it before it sends the rest. Records that someone else changed after you loaded them are skipped and listed; it stops on an error or a result it cannot confirm. Any batch of changes can be undone before you commit. You choose in the commit panel how many records one commit sends (all of them, or a number; 200 by default).
- Keep a record of what was approved. The commit panel saves a diff report (Excel): one row per changed cell with the old value, the new value, who changed it (the AI assistant or a person) and why. After a commit, it also holds the write log of that commit.
- Nothing leaves your PC except calls to your Maximo and your AI assistant. There is no MX Stage cloud and no telemetry. There are two optional exceptions: if you turn on Settings → Updates → Update automatically (off by default), the bridge asks GitHub once a day for the latest version number; and if you choose Settings → Demo → Download data and connect, it downloads fictional sample data once from
mxstage-demo.pages.dev(nothing is sent). Installs made with the setup script then update themselves while no work is open (the new version from GitHub, its packages from the npm registry); the Claude Desktop extension downloads the new file from GitHub and checks it for you to install. A small local process (the *bridge*) serves the work screen, talks MCP to your AI assistant and relays requests to Maximo, so Maximo needs no CORS settings. Your Maximo API key is saved on this PC by the bridge, encrypted with Windows data protection (DPAPI), and is not sent to the browser (on macOS, untested, the Keychain protects it; on Linux, a file only your user can read). Every window of the work screen — the installed app, a browser tab, or the browser inside your AI assistant — connects automatically, even after a restart, and you can switch between saved Maximo environments in Settings. (You can also connect without saving; the key then stays only in the tab's memory.) - Teach it your procedures. Save a procedure worked out in a conversation as a *Skill*; the AI follows it next time.
- Keep the customer's way of writing. *Customer dictionaries* hold how a customer writes names, the words that mean a failure code, and how values such as maker or department names are written. The AI adds entries and the work screen applies them as rules (matching, codes from text, value replacement), so the next month's files need less reading. You check and save them in the work screen. Each connection keeps its dictionaries and Skills in its own folder (Documents\MX Stage\ by default); choose a shared folder (OneDrive, SharePoint, a network folder…) in Settings → Skills & dictionaries, and saved entries reach every member through your own file sync. MX Stage itself sends them nowhere.
Why not call the Maximo REST API from the AI directly?
Several Maximo MCP servers let the AI call the REST API. That is fine for reading, but risky for bulk changes: you cannot see what the AI is about to write, and a wrong guess goes straight into production. MX Stage puts a staging sheet, a diff, an undo and a human approval in between.
Supported AI assistants
| Assistant | Status |
|---|---|
| Claude Desktop (chat and Code tab) | Supported |
| ChatGPT desktop — Codex or Work mode (Chat mode cannot use local MCP servers) | Supported |
| Antigravity (2.0, IDE) | Supported (registered by the installer) |
| IBM Bob | Supported (registered by the installer) |
| Claude Code, Codex CLI and IDE extension | Work, not officially supported |
Requirements
- Windows (macOS and Linux are untested)
- For the installer: Node.js 22.6 or later and Git (the installer offers to install them). The Claude Desktop extension needs neither; the npm package (
npx @mxstage/mxstage) needs Node.js 20 or later. - Chrome or Edge
- IBM Maximo or Maximo Application Suite (Manage) with the JSON API at
/maximo/apiand an API key
Install
Claude Desktop (simplest)
- Download
mxstage- .mcpbfrom the latest release. - In Claude Desktop, open Settings → Extensions, then Advanced settings → Install Extension… and choose the file (or double-click the file, or drag it onto the Extensions page). Choose Install.
- Open
http://127.0.0.1:8788/appin Chrome or Edge and enter your Maximo URL and API key in Settings. Never paste the API key into the chat. - In a new chat, ask "show the MX Stage status".
The extension needs no Node.js or Git: Claude Desktop runs it. To use MX Stage from other assistants as well (ChatGPT desktop, IBM Bob, Claude Code in the terminal, Antigravity), use the installer below; it detects the extension and does not register MX Stage twice.
Try without Maximo
No Maximo at hand? MX Stage includes a demo: a fictional Maximo with three waste incineration plants (assets, locations, about 52,000 work orders, inventory and the data quality problems migrations leave behind), in Japanese or English, running inside the bridge on your PC.
- Install MX Stage (above), then open
http://127.0.0.1:8788/settings#demo. - Choose Download data and connect. MX Stage downloads about 10 MB of fictional data (no programs) once from
mxstage-demo.pages.devand checks it against a SHA-256 built into this version. - Ask your AI assistant, for example "Check the data quality of the operating assets at North Clean Center and list the problems with counts."
Commits change only the copy on your PC (Reset to the initial state puts it back), and no license is needed. The demo takes about 0.5 GB of memory while in use. Five sample Excel files to merge with Maximo (purchase orders without work order numbers, a legacy equipment register, daily work reports printed as A4 forms (a repair log table in the English data), an equipment register ahead of Maximo, a star chart) are in the same tab. See docs/demo.en.md.
Claude plugin (Skills)
The mxstage plugin in plugin/ adds two Skills to Claude (planning a safe bulk correction of Maximo data; explaining, installing and troubleshooting MX Stage) and, in Claude Code and Cowork, starts the MX Stage server with npx @mxstage/mxstage (Node.js 20 or later). Chat does not start local servers, so use the extension there. In Claude Code: /plugin marketplace add mxstage/mxstage, then /plugin install mxstage@mxstage.
Other assistants (installer)
Ask Claude Code (or the Code tab of Claude Desktop) to install it: give it this repository's URL and say "install this". It follows Installation steps for Claude Code below and asks before running each command.
To install by hand, clone this repository to %USERPROFILE%\mxstage and run node scripts/setup-local.mjs in it.
One install registers MX Stage with every supported assistant found on this PC. Then restart the assistant you use, open mxstage on the desktop (or http://127.0.0.1:8788/app), and enter your Maximo URL and API key in Settings. Never paste the API key into the chat.
In Settings, also choose for each Maximo whether it is production or test. Committing to production needs a license (see License); everything else is free.
MCP clients that start servers with npm can also run the server alone with npx -y @mxstage/mxstage (Node.js 20 or later; the Skills and shortcuts come only with the installer).
To update, say "update MX Stage"; to remove it, say "uninstall MX Stage" (see Uninstall).
Where it is registered
~ is %USERPROFILE%. The installer only touches assistants that are installed (Claude Desktop: its settings folder exists; Antigravity: ~\.gemini; Codex: ~\.codex; IBM Bob: ~\.bob), backs up every file before changing it (to ~\.config\mxstage\backup\), and never changes other servers' settings. node scripts/setup-local.mjs --status shows what is registered without changing anything. To leave an assistant out, add --no-claude-desktop, --no-antigravity, --no-codex or --no-bob (--no-skills copies no Skills). --claude-code registers Claude Code even when the extension is enabled.
| Assistant | MCP server | Skills | After installing or updating |
|---|---|---|---|
| Claude Code (CLI) | mcpServers.mxstage in ~\.claude.json. Not registered when the MX Stage extension (.mcpb) is enabled in Claude Desktop, because the Code tab already gets MX Stage from the extension and would show every tool twice (an entry the installer added earlier is removed). Add --claude-code to register anyway for Claude Code in a terminal (remembered for later runs) | ~\.claude\skills\ \SKILL.md | Restart Claude Code |
| Claude Desktop (Code tab) | Same as Claude Code, or the extension (.mcpb) when it is enabled | Same as Claude Code | Quit Claude Desktop from the system tray and open it again |
| Claude Desktop (chat) | mcpServers.mxstage in %APPDATA%\Claude\claude_desktop_config.json. Microsoft Store version: also in %LOCALAPPDATA%\Packages\Claude_ \LocalCache\Roaming\Claude\claude_desktop_config.json when that folder exists. Not registered when the MX Stage extension (.mcpb) is installed and enabled in Claude Desktop (an entry the installer added earlier is removed) | Not copied (see Skills) | Quit from the system tray (closing the window is not enough) and open again |
| Antigravity | mcpServers.mxstage in ~\.gemini\config\mcp_config.json | ~\.gemini\config\skills\ \SKILL.md | Start a new conversation |
| Codex (ChatGPT desktop, CLI, IDE extension) | [mcp_servers.mxstage] in ~\.codex\config.toml | ~\.agents\skills\ \SKILL.md | Restart Codex |
| IBM Bob | mcpServers.mxstage in ~\.bob\settings\mcp.json | ~\.bob\skills\ \SKILL.md | Restart IBM Bob |
The installer copie
Solo se muestra el principio; lee el resto en el repo