AstroFabric
Servidor MCP de AstroFabric
AstroFabric es una plataforma de IA agéntica: agentes autónomos que llevan el trabajo de crecimiento, ingresos y operaciones digitales desde el objetivo hasta el resultado. Este servidor MCP pone toda la plataforma detrás de tu editor: dale un objetivo abierto en lenguaje natural y planifica y ejecuta la misión: inteligencia de mercado y competitiva, auditorías técnicas y del sitio, listas verificadas de prospectos a partir de señales de compra en vivo, planes de outbound, contenido y demanda, creatividades terminadas desde imágenes y vídeo hasta páginas diseñadas, y entrega en más de 100 aplicaciones conectadas, devolviendo el resultado final con la evidencia que lo respalda.
Este repositorio es tanto el conector para el endpoint MCP alojado de AstroFabric (añade la URL remota con tu clave de API y empieza a trabajar) como un servidor stdio ejecutable para clientes que prefieren un proceso local; consulta Ejecutar localmente.
Sitio web: https://www.astrofabric.ai
Endpoint:
https://www.astrofabric.ai/api/mcp(HTTP Streamable)Documentación: https://www.astrofabric.ai/docs
Requisitos previos
Un espacio de trabajo de AstroFabric en un plan de pago: regístrate en https://www.astrofabric.ai
Una clave de API: en la consola de AstroFabric, abre API keys → Create key y copia el token
ek_live_...(se muestra una sola vez).
Related MCP server: Agentic Control Framework (ACF)
Configuración
Todos los clientes usan el mismo endpoint (https://www.astrofabric.ai/api/mcp) y la misma clave. Elige el tuyo:
claude mcp add --transport http astrofabric https://www.astrofabric.ai/api/mcp --header "Authorization: Bearer ek_live_YOUR_KEY_HERE"Añade un conector personalizado con la URL simple e inicia sesión cuando se te pida: el servidor te guía por el inicio de sesión de AstroFabric y entrega al cliente una clave con ámbito que puedes revocar desde la consola:
https://www.astrofabric.ai/api/mcpAñade a .cursor/mcp.json (proyecto) o ~/.cursor/mcp.json (global):
{
"mcpServers": {
"astrofabric": {
"url": "https://www.astrofabric.ai/api/mcp",
"headers": { "Authorization": "Bearer ek_live_YOUR_KEY_HERE" }
}
}
}Añade a cline_mcp_settings.json (Cline → MCP Servers → Configure, o Remote Servers → añadir por URL):
{
"mcpServers": {
"astrofabric": {
"type": "streamableHttp",
"url": "https://www.astrofabric.ai/api/mcp",
"headers": {
"Authorization": "Bearer ek_live_YOUR_KEY_HERE"
},
"disabled": false,
"autoApprove": []
}
}
}Añade a .vscode/mcp.json:
{
"servers": {
"astrofabric": {
"type": "http",
"url": "https://www.astrofabric.ai/api/mcp",
"headers": { "Authorization": "Bearer ek_live_YOUR_KEY_HERE" }
}
}
}Añade a ~/.codex/config.toml:
[mcp_servers.astrofabric]
url = "https://www.astrofabric.ai/api/mcp"
http_headers = { "Authorization" = "Bearer ek_live_YOUR_KEY_HERE" }Añade a ~/.gemini/settings.json bajo mcpServers:
{
"astrofabric": {
"httpUrl": "https://www.astrofabric.ai/api/mcp",
"headers": { "Authorization": "Bearer ek_live_YOUR_KEY_HERE" }
}
}Añade a ~/.codeium/windsurf/mcp_config.json:
{
"mcpServers": {
"astrofabric": {
"serverUrl": "https://www.astrofabric.ai/api/mcp",
"headers": { "Authorization": "Bearer ek_live_YOUR_KEY_HERE" }
}
}
}Atajo: con la CLI instalada, astrofabric connect <client> imprime la configuración de tu cliente con tu clave ya rellenada.
Si tu cliente no puede enviar cabeceras, la clave puede ir en la URL: https://www.astrofabric.ai/api/mcp?key=ek_live_YOUR_KEY_HERE: trata toda esa URL como un secreto.
Dos superficies de herramientas
Predeterminada (la URL anterior): una herramienta,
mission_agent: el ejecutor de misiones autónomo. Ideal para editores: una entrada en tu lista de herramientas, toda la plataforma detrás.Catálogo completo: añade
?tools=alla la URL para exponer también todas las herramientas individuales de datos y ejecución (inteligencia de SEO y palabras clave, bibliotecas de anuncios, búsqueda de intención de compra, búsqueda y verificación de correos electrónicos, generación de creatividades y más), además de cualquier servidor MCP que tu espacio de trabajo tenga conectado.
Cómo usarlo
Pide resultados, no pasos:
"Crea una lista de 40 empresas que muestren intención de compra en robótica de almacenes, verifica los correos electrónicos y redacta un primer contacto para cada una."
"Haz un análisis de competitor.com: anuncios, palabras clave, contrataciones, posicionamiento, y dame las tres jugadas para ganar terreno."
Los resultados comienzan con una línea [thread:<id>]. Pasa ese id como thread_id en los seguimientos ("verifica esos correos", "dales formato para LinkedIn Ads") y el agente continúa con memoria completa de todo lo ya preguntado y entregado. Puede responder con una pregunta aclaratoria; respóndela de la misma manera.
Ejecutar localmente (stdio)
Este repositorio también es un servidor MCP stdio ejecutable: la misma herramienta mission_agent, ejecutando misiones a través de la plataforma AstroFabric con tu clave. Para clientes que prefieren un proceso local en lugar de una URL remota:
{
"mcpServers": {
"astrofabric": {
"command": "npx",
"args": ["-y", "github:sam1siam/astrofabric-mcp"],
"env": { "ASTROFABRIC_API_KEY": "ek_live_YOUR_KEY_HERE" }
}
}
}O con Docker: docker build -t astrofabric-mcp . && docker run -i --rm -e ASTROFABRIC_API_KEY=ek_live_... astrofabric-mcp
¿Prefieres un terminal? La CLI
La misma plataforma se distribuye como una CLI sin dependencias (astrofabric en npm):
npm install -g astrofabric
astrofabric login --key ek_live_...
astrofabric run "Audit example.com and list the top 5 fixes"En un agente de codificación (Claude Code, Codex, Gemini CLI), toda la configuración es un solo prompt: "Instálame la CLI de AstroFabric y luego ejecuta astrofabric docs para aprender a usarla." Y astrofabric connect cursor (o claude-code, codex, gemini, vscode, windsurf) imprime la configuración MCP lista para pegar por cliente.
Notas
Las misiones se facturan contra los créditos de tu espacio de trabajo; la página Usage de la consola muestra cada ejecución.
Las escrituras en sistemas conectados respetan la configuración de aprobación de tu espacio de trabajo: las escrituras autónomas pueden estar sujetas a una cola de aprobación humana.
También se admite OAuth 2.1 (registro dinámico de clientes + PKCE) mediante descubrimiento estándar, para clientes que prefieren iniciar sesión en lugar de claves.
Licencia
El contenido de este repositorio de conectores tiene licencia MIT. La plataforma AstroFabric es un servicio comercial.
Available Tools
1 toolmission_agentMission Agent (autonomous missions)A
Give AstroFabric an open-ended objective across growth, revenue or digital operations in plain language and it plans and executes the whole mission autonomously: research markets, audit sites, analyze competitors, build prospect lists from buyer intent and signals, verify emails, draft outbound, produce creative, run data work in the code sandbox, deliver into connected apps, and more. The result starts with a [thread:] line - pass that id as thread_id on follow-ups ("verify those emails", "format them for LinkedIn Ads") and the agent remembers everything already asked and delivered. It may reply with a clarifying question; answer it the same way. Prefer this over the individual tools for anything multi-step.
| Name | Required | Description | Default |
|---|---|---|---|
| objective | Yes | What to accomplish, e.g. "Build a list of 100 companies showing high interest in building insulation" | |
| thread_id | No | The thread id from a previous result, to continue that mission with full memory |
TDQS
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
The description goes far beyond the annotations by explaining autonomous planning/execution, the [thread:<id>] result prefix, memory across follow-ups, and the possibility of a clarifying question. It also mentions capabilities like running data work in the code sandbox and delivering into connected apps, which gives an agent a concrete sense of side effects. No contradiction with annotations was mentioned.
Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.
Is the description appropriately sized, front-loaded, and free of redundancy?
The description is dense but every sentence adds new information: core functionality, key output format, continuation pattern, clarification behavior, and usage preference. The front-loaded sentence explains scope first, followed by operational details. It is aligned with the best vertex for high-functioning agents.
Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.
Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?
Given the tool is an autonomous agent with no sibling tools and no output schema, the description surprisingly covers everything essential: what the objective should be, what outcomes can be expected, how to start, how to continue, how to handle clarifying questions, and when to prefer it. The emptiness of a formal output schema is compensated by the explicit mention of the thread line at the beginning. This is a high-quality, self-sufficient description.
Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.
Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?
The input schema already covers 100% of parameters with meaningful descriptions and an example for objective. The description reinforces thread_id usage in the continuation context but does not materially add new semantic details beyond the schema. This matches the baseline where the schema carries the parameter-load.
Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.
Does the description clearly state what the tool does and how it differs from similar tools?
The description clearly states that the tool takes a plain-language objective and autonomously plans/executes a multi-step mission across growth, revenue, and digital operations. It lists concrete actions (research, audits, competitor analysis, email verification) and explicitly contrasts itself with individual tools. This goes well beyond a vague restatement of the name.
Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.
Does the description explain when to use this tool, when not to, or what alternatives exist?
It explicitly says 'Prefer this over the individual tools for anything multi-step', which specifies when to use it and strongly implies when not to use it (single-step operations). It also documents the follow-up workflow using thread_id and how to respond to clarifying questions. This is strong, actionable usage direction.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Tool Schema Changelog
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
1 tool update
v1.1.0- First observed
mission_agent
TDQS
Scored across 1 tool
The server exposes exactly one tool, so there is no possibility of confusing it with another. All tool selection is unambiguous by definition.
The lone tool name 'mission_agent' is clear and internally consistent, but the pattern cannot be validated across a set. No naming conflicts or inconsistencies exist.
A single tool is borderline for a server that describes a broad range of capabilities. The one mission_agent can be a legitimate monolithic entry point, but the count feels thin given the scope.
The mission_agent handles open-ended missions and follow-ups via thread_id, but there are no explicit controls for listing, canceling, or monitoring missions. The description hints at individual granular tools that are absent from this server surface.
Maintenance
Related MCP Connectors
Orbit is an MCP server that gives your AI agents a 4000+ tools, a shared memory brain and a live dashboard OS. Every agent reads active decisions, logs outputs to a searchable vault, and reports status in real time. 2-minute setup via MCP; works across Claude, ChatGPT, Codex, Cursor, Gemini, and Manus. Finally Agents work, agents that grow your business and bring you customers. https://orbitagents.xyz
Connect, monitor, and control AI agents — tasks, approvals, schedules, and governance.
MCP server for your apps' tools and custom tools, plus hosted AI agents and approval-gated workflows
Agent-native security, trust, reliability, data and procurement tools for AI workflows.
Related MCP Servers
- FlicenseNot gradedqualityCmaintenanceEnterprise-grade MCP server with multi-agent system for business AI transformation across finance, healthcare, retail, and other domains. Provides specialized AI agents for data analysis, API execution, business validation, and report generation with real-time monitoring and observability.-
- AlicenseDqualityDmaintenanceAI-native orchestration layer with 80+ tools for task management, code editing, browser automation, terminal control, and persistent memory across CLI, local MCP, and cloud deployments.694 npm31ISC
- AlicenseAqualityBmaintenanceAI Agent Mission Control — 200+ MCP tools across 31 domains. Manage agents, experiments, workflows, crews, skills, tools, credentials, approvals, signals, budgets, marketplace, knowledge bases, chatbots, and more. Self-hosted, open-source (AGPL-3.0). Supports stdio + Streamable HTTP/SSE with OAuth 2.0 auth.3477AGPL 3.0
- AlicenseNot gradedqualityDmaintenanceProvides AI agents with operational customer context, including typed revenue objects, persistent state, scoped tools, and human-in-the-loop handoffs through MCP, REST, and CLI.15 npm13Apache 2.0