Spec Driven Development
MCP Server that facilitates spec-driven development workflows, not just Vibe Coding.
Descripción del creador (original en inglés)
A shopper's agent searches stores, compares prices and buys with the shopper's approval.
Estos tres nunca se combinan en una sola puntuación: las opiniones, las mediciones y la popularidad son cosas distintas.
Todavía no escribimos un resumen sencillo para este. Abajo están los datos que leímos del registro.
Elige la app de IA que usas y copia la configuración.
Necesitas
¿Qué app usas?
Elige tu app arriba. La recordamos, así que la próxima vez el botón dirá “Instalar en …”.
Abre Claude Desktop → Settings → Developer → Edit Config y pega esto en claude_desktop_config.json
{
"mcpServers": {
"portage": {
"command": "docker",
"args": [
"run",
"-i",
"--rm",
"ghcr.io/tomtom87/portage-mcp:0.15.0"
]
}
}
}Ejecuta esto en tu terminal
claude mcp add portage -- docker run -i --rm ghcr.io/tomtom87/portage-mcp:0.15.0Pégalo en .cursor/mcp.json de tu proyecto, o en ~/.cursor/mcp.json
{
"mcpServers": {
"portage": {
"command": "docker",
"args": [
"run",
"-i",
"--rm",
"ghcr.io/tomtom87/portage-mcp:0.15.0"
]
}
}
}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.
Pégalo en .vscode/mcp.json de tu proyecto
{
"servers": {
"portage": {
"type": "stdio",
"command": "docker",
"args": [
"run",
"-i",
"--rm",
"ghcr.io/tomtom87/portage-mcp:0.15.0"
]
}
}
}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.
Se ejecuta en tu computadora; la lista de herramientas aparece en tu app de IA después de instalarlo.
Revisamos información que cualquiera puede ver; no hemos leído el código. Las revisiones que no aplican no se cuentan. · Revisado el 7 nov 2026
El nombre en el registro se verificó con la cuenta de GitHub tomtom87, dueña del repo
Todavía no hemos leído este repo en GitHub
Todavía no hemos leído este repo en GitHub
Este se ejecuta en tu computadora; no hay endpoint en internet
Se ejecuta en tu computadora y usa ajustes locales en su lugar
Se ejecuta en tu computadora; solo podemos leer listas de herramientas de endpoints en internet
Comparado con todos los servidores activos de esta categoría en el registro. Las estrellas solo cuentan cuando quien lo publica es dueño del repo.
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.
El registro no indica ninguna clave obligatoria, aunque algunos servidores te piden iniciar sesión la primera vez que los usas.
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).
El repo de GitHub cambió por última vez –, y la última versión (0.18.0) llegó al registro hace 16 días.
Las apps compatibles con MCP, como Claude Desktop, Claude Code, Cursor y VS Code. También necesitas Docker en tu computadora.
En la misma categoría: Spec Driven Development, Remote Desktop Commander, Claude Checkpoint MCP
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
MCP Server that facilitates spec-driven development workflows, not just Vibe Coding.
Remote Desktop Commander: use files, terminal and processes on your own computers from any AI app.
MCP server that checkpoints Claude Code session context to survive crashes and token limits
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.
Portage lets an AI agent find and buy things from real online stores for you, and you approve every payment. It ships as a command-line tool (portage), a Claude Code plugin (buy) that drives it, and Ruby gems that let any store serve the same open protocols (MCP and UCP) to shopping agents. It is for people who want an agent to shop for them, developers building shopping agents, and merchants who want agents to buy from their store.
Status: pre-1.0. APIs may still change between minor versions. Latest release set: 0.22.0 (changelog).
Two ways in: let Claude shop for you with the buy plugin, or run the portage CLI yourself. Both use the same CLI and take about five minutes.
buy pluginThe plugin teaches Claude Code to shop through the portage CLI, so install the CLI first.
portage-cli 0.9.0 or newer):bash brew install tomtom87/portage/portage
Or, on Ruby 3.2 or newer, gem install portage-cli. Homebrew bundles every adapter. With RubyGems, also install portage-ucp-webmcp 0.2.0 or newer if you want the Portage browser profile and checkout autofill.
text /plugin marketplace add tomtom87/Portage /plugin install buy@portage
Or from a shell:
bash claude plugin marketplace add tomtom87/Portage claude plugin install buy@portage
/buy a burton snowboard under $600, or just ask in plain words. The skill is listed as buy:buy. Claude checks your setup, shows real offers and a dry-run total, and waits for your yes before any purchase.To update or remove it:
claude plugin marketplace update portage
claude plugin update buy@portage # then restart Claude Code
claude plugin uninstall buy@portage claude plugin marketplace remove portage
A listing in the Claude plugin directory is coming. Until then, this marketplace is the install route.
The buy skill is a plain SKILL.md file, so any agent that loads skills and can run commands on your machine can use it. Install the CLI and run portage setup first, as above. The plugin carries a second skill, shop-research: read-only, for price, stock, store and order questions with no purchase in mind, and it hands over to buy to buy. Install it next to buy the same way.
~/.agents/ and generates each agent's plugin or skill files.bash npx @sentry/dotagents add tomtom87/Portage
Refresh it with npx @sentry/dotagents install. Remove it with npx @sentry/dotagents remove buy. If your agents.toml only allows trusted sources, run npx @sentry/dotagents trust add tomtom87/Portage first.
plugins/buy/skills/buy/ folder into the agent's skills directory, and plugins/buy/skills/shop-research/ next to it. With dotagents, declare them in ~/.agents/agents.toml and run npx @sentry/dotagents install to get them in ~/.agents/skills/buy and ~/.agents/skills/shop-research:`toml [[skills]] name = "buy" source = "tomtom87/Portage" path = "plugins/buy/skills/buy"
[[skills]] name = "shop-research" source = "tomtom87/Portage" path = "plugins/buy/skills/shop-research" `
openclaw skills install @tomtom87/portage-buy puts it in your active OpenClaw workspace, and clawhub install @tomtom87/portage-buy puts it in ./skills under the current directory (ClawHub docs). Then run portage setup. The skill's frontmatter carries OpenClaw's metadata.openclaw block (needs the portage binary, names every env var it reads as optional, brew install spec), so OpenClaw gates the skill until portage is installed. To skip ClawHub, use the plain-skill route above: OpenClaw reads personal skills from ~/.agents/skills and managed ones from ~/.openclaw/skills (OpenClaw skills docs), so copy or symlink plugins/buy/skills/buy/ into either. shop-research is on ClawHub too, as @tomtom87/portage-shop-research (install it the same way), or copy or symlink plugins/buy/skills/shop-research/. Its frontmatter has the same metadata.openclaw shape, with only the search and retailer keys it names. ClawHub republishes skills under MIT-0; this repo stays MIT.openclaw plugins install clawhub:tomtom87/portage, after the CLI (portage-cli 0.12.0 or newer) and portage setup. It gives the agent portage_* tools over the CLI and bundles buy, shop-research and a short OpenClaw skill that maps each buy step to its tool. The read-only tools are on by default; the ones that price, approve, buy or hand off are opt-in in OpenClaw's tool settings. Spending policy and approval stay in Portage, and no tool can pass --yes for a URL or loosen your limits. See the plugin README.omarchy-mise-install gem:portage-cli portage, which writes a ~/.local/bin/portage wrapper the way Omarchy installs claude, codex and gh. Plain mise (mise use -g gem:portage-cli) or Homebrew on Linux (brew install tomtom87/portage/portage) also work. Then run portage setup. A stock Omarchy needs sudo pacman -S --needed make first: Ruby 3.4 builds bigdecimal natively, gcc is already there through clang, and only make is missing. For the skills, link plugins/buy/skills/buy/ and plugins/buy/skills/shop-research/ into the skills directory of the agent you run. Omarchy's own provisioning links into ~/.agents/skills (OpenClaw, and the dotagents route), ~/.claude/skills (Claude Code; the plugin route above is preferred) and ~/.codex/skills.portage on your machine, so they can't use the skill. Use the vendor's coding agent or CLI instead, if it loads skills.portage mcp serves the same search, quote, approve and buy tools over MCP on stdio, for any client that runs local servers. Install the CLI (brew install tomtom87/portage/portage or gem install portage-cli, 0.14.0 or newer) and run portage setup in a terminal first. Spending policy and approval stay in Portage: the agent can search and price, but a purchase waits for your approval.
claude_desktop_config.json (Settings, Developer, Edit Config) and restart. Claude Desktop doesn't inherit your shell PATH, so give the absolute path from which portage (typically /opt/homebrew/bin/portage or /usr/local/bin/portage):json { "mcpServers": { "portage": { "command": "/opt/homebrew/bin/portage", "args": ["mcp"] } } }
claude mcp add --scope user portage -- portage mcp (drop --scope user to keep it to the current project).docker run -i --rm -v "$HOME/.portage:/root/.portage" ghcr.io/tomtom87/portage-mcp:0.14.0 runs the same server from the registry image. A container can't open your browser, so a hand-off prints the checkout link instead.Install the CLI as in step 1 above, run portage setup, then:
portage find --query "burton snowboards" --json
portage buy "burton snowboards" --max-price 600 --dry-run --json
find works with no keys. Its default search, DuckDuckGo's keyless API, resolves brand and store names ("burton snowboards") but not open-ended queries ("waterproof hiking boots"). For those, add a Brave or Google search key: portage setup asks, and search backends has the details. buy without a store URL lists offers and, in a terminal, lets you pick one. --dry-run shows the real total and never charges.
With a store URL, buy goes straight to that store's /.well-known/ucp manifest:
portage buy https://some-ucp-store.example --query "hoodie" --dry-run --json
To buy for real, enroll a payment method (portage payment enroll ), set spending caps (portage policy set), then drop --dry-run and add --yes. buy checks out only through a store's published UCP endpoint, or a platform API you already hold credentials for (your own store). Everything else ends in a hand-off. It never scrapes. The CLI tutorial walks through all of this, including what to do when search comes back empty.
portage setup saves to ~/.portage/.env (chmod 600), and .env.example lists every variable. Only ~/.portage/.env loads automatically, never a .env in the current directory, so a cloned repo can't quietly redirect your p
Solo se muestra el principio; lee el resto en el repo