MCP server de AgentFirstStack Analytics
Commit 11 nov 2026Permite a los agentes de IA configurar y gestionar un stack completo de go-to-market que incluye analítica sin cookies, gestión de etiquetas, bandejas de entrada de correo electrónico y grabaciones de sesiones.
También conocido como StartGlobal-AI/agentfirststack, agentfirststack
- Stars
- 1
- Herramientas
- —
- Descargas, 30 días
- —
- Última versión
- 0.1.1
- Último commit
- 11 nov 2026
- Licencia
- AGPL-3.0
¿Qué es AgentFirstStack Analytics MCP?
AgentFirstStack Analytics es un MCP server. Permite a los agentes de IA configurar y gestionar un stack completo de go-to-market que incluye analítica sin cookies, gestión de etiquetas, bandejas de entrada de correo electrónico y grabaciones de sesiones.
Resumen generado a partir del README del proyecto
Acerca del MCP server de AgentFirstStack Analytics
AgentFirstStack Analytics proporciona una infraestructura completa de go-to-market diseñada específicamente para que los agentes de IA la desplieguen sin necesidad de interacción manual con un dashboard. Sustituye los flujos de configuración tradicionales que requieren formularios de registro y correos de verificación por una única solicitud de API. El sistema incluye analítica web sin cookies, un gestor de etiquetas para el seguimiento de conversiones en el lado del servidor, bandejas de entrada de correo electrónico y herramientas de grabación de sesiones.
Los desarrolladores pueden utilizar esto mediante una versión alojada o auto-alojándola en Cloudflare Workers. Al usar el servicio alojado, un agente puede configurarse a través de Claude Code, Cursor u otros clientes utilizando herramientas MCP. El proceso de configuración implica mapear las páginas y acciones del sitio, y luego verificar el despliegue mediante tráfico de prueba automatizado. La plataforma admite la importación de historial desde herramientas como GA4 o Plausible y permite que los humanos reclamen proyectos mediante un enlace de inicio de sesión de Google para visualizar los datos en un dashboard.
Qué puedes hacer con él
- 1Seguimiento de conversiones en el lado del servidorConfigura un gestor de etiquetas para dirigir eventos de conversión a múltiples plataformas publicitarias simultáneamente utilizando un único ID de evento.
- 2Verificación de correo electrónico automatizadaUtiliza el servicio de correo para recibir y extraer programáticamente códigos de verificación de los mensajes entrantes.
- 3Grabación y análisis de sesionesCaptura las interacciones de los usuarios para identificar errores técnicos o frustración del usuario mediante mapas de calor y transcripciones.
Escrito a partir del README del proyecto. Cada caso de uso se verifica con una cita del mismo.
Cómo instalar AgentFirstStack Analytics MCP
Formatos verificados el 10 nov 2026Instalar AgentFirstStack Analytics MCP en Claude Code
$ claude mcp add --transport http agentfirststack https://analytics.agentfirststack.com/mcpInstalar AgentFirstStack Analytics MCP en Cursor
~/.cursor/mcp.json
{
"mcpServers": {
"agentfirststack": {
"url": "https://analytics.agentfirststack.com/mcp"
}
}
}Instalar AgentFirstStack Analytics MCP en VS Code
.vscode/mcp.json
{
"servers": {
"agentfirststack": {
"type": "http",
"url": "https://analytics.agentfirststack.com/mcp"
}
}
}Instalar AgentFirstStack Analytics MCP en Codex
~/.codex/config.toml
[mcp_servers.agentfirststack] url = "https://analytics.agentfirststack.com/mcp"
Instalar AgentFirstStack Analytics MCP en Gemini CLI
~/.gemini/settings.json
{
"mcpServers": {
"agentfirststack": {
"httpUrl": "https://analytics.agentfirststack.com/mcp"
}
}
}Configuración
CLOUDFLARE_API_TOKEN
| Nombre | Descripción | Tipo |
|---|---|---|
| CLOUDFLARE_API_TOKEN | — |
Verificaciones automáticas
Ejecutado 14 nov 2026Informe completo →Aún no verificado. Un servidor necesita una puntuación de 80, ningún hallazgo grave y suficientes datos para que podamos revisarlo. Cómo funcionan las verificaciones
- Repositorio de origen vinculado y accesiblestartglobal-ai/agentfirststack is public and reachablegithub.com/startglobal-ai/agentfirststack ↗Aprobado
- El paquete cita este repositorio como fuenteNo verificado
- Procedencia de la compilación (atestación Sigstore)No verificado
- Licencia de código abierto (OSI)AGPL-3.0spdx.org/licenses/AGPL-3.0.html ↗Aprobado
- Commit o lanzamiento en el último añoLast commit or release 2026-11-11github.com/startglobal-ai/agentfirststack ↗Aprobado
- OpenSSF ScorecardNo verificado
- El nombre del paquete no es similar a un paquete popularNo verificado
- Sin avisos críticos en el paquete ni en sus dependenciasNo verificado
- Sin instrucciones ocultas para el modelo0 tool descriptions and the README read; no override phrases, credential paths or hidden charactersAprobado
- Las herramientas destructivas están etiquetadasNo verificado
- El endpoint remoto responde al handshake MCPHTTPS endpoint answered 401 (sign-in required)Aprobado
README
Ver en GitHubEvery new website needs the same infrastructure: analytics, a tag manager with conversion tracking, and an email address to sign up for everything else. The usual tools all start with a signup form, a verification email and a dashboard. Coding agents can't get past those steps alone.
AgentFirstStack is that infrastructure, rebuilt so an agent can do the whole job:
- One request gets an account.
POST /v1/signupreturns a key that works on every service. - Everything is an API call and an MCP tool. Nothing is hidden behind a settings page.
- The manual is a text file. Each service serves
/llms.txtwith every endpoint, tool and field. - Every step can be verified. The agent checks its own work instead of guessing.
- Served from your own domain. Analytics, tags and replay load from a first-party domain (one CNAME, no server code) or a path proxy, so ad blockers don't drop your visitors.
- Humans optional. When it's done, the agent hands its person a claim link. They sign in with Google to see the project's analytics, tags, replay and mail in a dashboard, manage keys and teammates, connect Claude or ChatGPT over OAuth, and delete the project. We email them about their projects, and alert them when a site stops sending data, gets suspicious traffic (one network over its share or a sudden flood is set aside, out of reports and budgets, for review), or a project gets close to a fair-use limit (docs/accounts.md).
Services
| What it does | Instead of | |
|---|---|---|
| Analytics | Cookieless web analytics. Country, region and city from Cloudflare's edge; channels including AI assistants; UTM and ad click ids; goals, funnels, paths, attribution, realtime (a live stream within a second); revenue, subscriptions and MRR from Stripe, Polar, Paddle or LemonSqueezy; ad spend and ROAS; ecommerce items; AI crawler visits and an AI search report (which AI bots read which pages, which assistants send visitors, what to change); Core Web Vitals from real visitors; alerts and weekly reports; history imported from GA4, Plausible or any other tool, with a check that the numbers match; server-side events; read-only SQL | Google Analytics, Plausible |
| Tags | A tag manager and server-side conversion router. Define a conversion once; the browser pixel and the server both send it to GA4, Google Ads, Meta, TikTok, LinkedIn, Pinterest, Reddit, Snap or X with one event id, so each platform counts it once. A consent banner asks visitors from the EU, UK and Switzerland before ad pixels fire. Scan (it also finds the tools a site uses now), import a GTM container, validate, preview, test and publish | Google Tag Manager, server-side GTM |
Inboxes in one request, on agentfirstinbox.com or your own domain. wait_for_email returns the moment a message arrives, with verification codes and links already extracted. Read, reply, send, custom domains, webhooks, and forwarding to the project's people (each email or a daily digest) |
A Gmail account per project, AgentMail | |
| Replay | Session recordings masked in the browser by default. Sessions come back as readable transcripts; rage, dead and error clicks, heatmaps and JS errors as data; a share link opens a player for people | Hotjar, Microsoft Clarity |
Quick start
Hosted (free)
Copy the prompt from agentfirststack.com into Claude Code, Cursor,
Codex or any agent that can make HTTP requests. The agent follows the
full setup in one run: it maps your site's pages, main
actions and the tools it uses now, then asks you once where the inbox's email should go and whether
to bring your old tools' setup and history over. It creates an inbox, sets up analytics (goals, AI
crawler visits reported from your server, a weekly report for you, and the history of GA4,
Plausible or whatever you used before), a tag container with an event for each main action
(starting from your GTM container, if you have one), and replay; then it deploys, verifies each
part as a visitor with test traffic (POST /v1/setup/test: a test email, test events through analytics, tags and each ad
platform's test mode, then a browser test visit), runs the setup check (GET /v1/setup) until
nothing is missing, writes what it did to AGENTFIRSTSTACK.md (and a short block to AGENTS.md, so later sessions with
any agent pick up new features), and shows you a table of what's set up and what passed, with a
claim link. Your old tools stay until the setup check says the numbers match and you agree to
remove them. Pasting the same prompt again later starts from the
setup check and fills the gaps.
Or install it as an agent skill (skills.sh), so your agent has the setup steps on hand in every session:
npx skills add StartGlobal-AI/agentfirststack
Then ask it to set up AgentFirstStack. The skill is
skills/agentfirststack-setup/SKILL.md.
Or do it by hand:
curl -X POST https://agentfirststack.com/v1/signup
# { "api_key": "afs_...", "mcp": { ... }, "next_steps": [ ... ] }
curl -X POST https://analytics.agentfirststack.com/v1/sites \
-H "Authorization: Bearer $AGENTFIRSTSTACK_API_KEY" -H 'content-type: application/json' \
-d '{"domain":"example.com"}'
Add one MCP server for every service (tools prefixed analytics_, tags_, mail_, replay_):
claude mcp add --transport http afs https://agentfirststack.com/mcp --header "Authorization: Bearer $AGENTFIRSTSTACK_API_KEY"
Or one per service:
claude mcp add --transport http afs-analytics https://analytics.agentfirststack.com/mcp --header "Authorization: Bearer $AGENTFIRSTSTACK_API_KEY"
claude mcp add --transport http afs-tags https://tags.agentfirststack.com/mcp --header "Authorization: Bearer $AGENTFIRSTSTACK_API_KEY"
claude mcp add --transport http afs-mail https://mail.agentfirststack.com/mcp --header "Authorization: Bearer $AGENTFIRSTSTACK_API_KEY"
claude mcp add --transport http afs-replay https://replay.agentfirststack.com/mcp --header "Authorization: Bearer $AGENTFIRSTSTACK_API_KEY"
The hosted version has fair-use limits, listed in the manual.
Self-hosted
Everything runs on Cloudflare Workers in your own account:
git clone https://github.com/StartGlobal-AI/agentfirststack && cd agentfirststack
pnpm install
pnpm afs init mystack --domain example.com --account <cloudflare-account-id>
pnpm afs deploy mystack
CLOUDFLARE_API_TOKEN=... pnpm afs email mystack # inbound + outbound mail
pnpm afs check mystack
See docs/self-hosting.md for requirements, signup modes, limits and updates.
How it's built
TypeScript on Cloudflare Workers with Hono and zod. Storage:
- D1: configuration and the shared identity database.
- SQLite Durable Objects: hot per-entity data (a site's events, a container's visitors, an inbox's waiters).
- R2: mail blobs and replay recording chunks.
- Queues: retried delivery to ad platforms and webhooks.
- Email Routing and Email Sending: mail in and out.
site/ landing page, manual, signup, MCP gateway, dashboard agentfirststack.com
analytics/ web analytics analytics.agentfirststack.com
tags/ tag manager + conversion APIs tags.agentfirststack.com
mail/ inboxes, sending, domains, webhooks mail.agentfirststack.com
replay/ session replay, signals, heatmaps, share player replay.agentfirststack.com
shared/ @afs/shared: auth, signup, errors, MCP handler
scripts/ pnpm afs: the deployment tool
deploy/ deployment profiles (yours; the agentfirststack.com one is kept private)
docs/ architecture, self-hosting, operations, development
More in docs/architecture.md.
Development
pnpm install
pnpm typecheck && pnpm test # every package; integration tests run inside workerd
cd mail && pnpm db:migrate:local && pnpm dev
docs/development.md covers running services locally, tests, and adding endpoints, migrations and tag destinations. docs/operations.md covers releases, migrations, secrets, monitoring and abuse handling for a running deployment.
Contributing
Issues and pull requests are welcome. Read CONTRIBUTING.md first; it's short. Report security issues privately as described in SECURITY.md.
License
The services are AGPL-3.0; the browser scripts websites load (the analytics tracker, the tag manager runtime, the replay recorder and survey widget) are MIT, so adding them to a site puts no obligations on it. Self-hosting is free either way; if you offer a changed version as a service, publish your changes. Details: LICENSING.md. Built and hosted by StartGlobal.
Preguntas sobre AgentFirstStack Analytics MCP
¿Cómo agrego AgentFirstStack Analytics MCP a Claude Code?
Ejecuta claude mcp add --transport http agentfirststack https://analytics.agentfirststack.com/mcp en tu terminal. Añade --scope user para que esté disponible en todos los proyectos.
¿Cómo agrego AgentFirstStack Analytics MCP a Cursor?
Abre ~/.cursor/mcp.json y añade el servidor en mcpServers, usando la configuración de Cursor que se muestra en la sección de instalación de arriba.
¿Puedo alojarlo yo mismo?
Sí, todo el stack se puede auto-alojar ejecutándose en Cloudflare Workers en tu propia cuenta.
¿Es compatible con datos analíticos existentes?
Sí, puedes importar el historial de GA4, Plausible u otras herramientas, y el sistema incluye una comprobación para asegurar que los números coincidan.
¿Qué tipo de autenticación utiliza?
El sistema utiliza API keys para el acceso de agentes y admite Google OAuth para que los usuarios humanos accedan al dashboard.