CIA Diagnose

CIA Diagnose

Expert business diagnosis engine that analyzes companies across 11 dimensions and returns a Revenue Leak Score with prioritized, triple-option recommendations.

Category
Visit Server

README

CIA Diagnose v1.1.1

<!-- mcp-name: io.github.davidscoreal/cia-diagnose -->

Expert business diagnosis engine by CIA — Consultoría de Inteligencia Aplicada.

Analyzes any company across 11 dimensions and returns a Revenue Leak Score with prioritized, triple-option recommendations (paid tool / open-source / CIA service). Works from Claude, ChatGPT, Gemini, DeepSeek, Cursor — any LLM that speaks MCP.

📅 Book a free 30-min diagnosis review: https://cal.com/david-cia/diagnostico-ai ✉️ steban@univercityaiconsult.tech · 🌐 https://www.univercityaiconsult.tech/


🚀 For prospects — copy & paste this into any AI

You don't need to know what an "MCP" is. Open Claude, ChatGPT, or any AI assistant and paste this:

Quiero un diagnóstico profesional de mi negocio.

Usa la herramienta de diagnóstico de CIA (Consultoría de Inteligencia
Aplicada) — se instala con: pip install cia-diagnose

Pasos:
1. Pregúntame sobre mi empresa: industria, tamaño del equipo, qué
   software uso, qué dolores tengo, cómo manejo proveedores, flujo de
   caja, etc.
2. Cuando tengas suficiente contexto, corre el diagnóstico. Si te falta
   información, pregúntame primero — no adivines.
3. Muéstrame mi Business Health Score (0-100, más alto = más sano; bajo =
   área con mayor oportunidad) y las acciones prioritarias con 3 opciones
   cada una (herramienta paga, alternativa gratuita, servicio profesional
   de CIA con precio).
4. Si el diagnóstico muestra oportunidades, quiero agendar una llamada
   gratuita con CIA: https://cal.com/david-cia/diagnostico-ai — o
   escribir a steban@univercityaiconsult.tech

Más info: https://www.univercityaiconsult.tech/

If your AI can connect MCP servers, it will use CIA Diagnose automatically. If not, pip install cia-diagnose is the universal fallback any code-capable assistant can run.


Architecture — "LLM has EYES, MCP has BRAIN"

The LLM gathers data from the user (the eyes). CIA Diagnose applies CIA consulting expertise (the brain) — ICP detection, 11-dimension scoring, revenue-leak estimation, and triple-option recommendations. No scraping. No fixed questionnaire. Each diagnosis is unique.

The 9 tools (the full value journey)

Tool Description
quick_scan FREE instant 3-problem scan. No context needed — hook the client.
business_diagnose Full 11-dimension analysis + Revenue Leak Score + triple-option actions. Forwards the lead to CIA.
list_industries Available industry benchmarks.
tools_recommend Free & open-source tools per weak dimension.
action_plan 30 / 60 / 90-day prioritized roadmap.
roi_projector ROI calculator — DIY vs CIA-guided vs full implementation.
case_studies Before/after transformation stories by industry.
contact_cia Services catalog, pricing, and booking link.
export_report Shareable structured report for the client's team.

The 11 dimensions

finanzas, comercial, operaciones, equipo, tecnologia, marketing, clientes, proveedores, legal, estrategia, marketing_digital.

Scoring — Business Health Score (0-100, higher = healthier)

The headline health_score is inverted from a leak score: high = healthy/strong, low = the worst area / biggest revenue-leak opportunity. A 100 doesn't mean "done" — it means that area is taking the business into new leagues / growth, which qualifies it even better for CIA. The output also returns score_meaning, growth_mode (true when most areas are strong — the conversation continues toward scaling), and guidance for the calling LLM (ask more vs. present + 3 options). revenue_leak_score is kept as a backward-compatible alias (same value).

Triple Option + curated tool registry

Every recommendation shows three paths: the best paid tool, the best open-source alternative, and the CIA professional service — with prices. The free/OSS/paid picks come from CIA's curated registry (tools_registry/, by area and per-industry), refreshed weekly. Each tool carries why_best. Pass industry to tools_recommend for industry-specific picks. See TOOLS_REGISTRY.md.

Quick Start

Install

pip install cia-diagnose      # or: uvx cia-diagnose

Run (stdio — local)

cia-diagnose

Run (HTTP — remote)

CIA_TRANSPORT=streamable-http cia-diagnose --transport streamable-http --port 3792

HTTP mode also serves:

  • GET /report/{session_id} — visual HTML diagnosis report (gauge + breakdown + actions).
  • GET /export/{session_id}?format=csv|json — export the diagnosis for Sheets / Tabularis.
  • GET /healthz — health check.

Claude Desktop config

{
  "mcpServers": {
    "cia-diagnose": { "command": "uvx", "args": ["cia-diagnose"] }
  }
}

Remote (Streamable HTTP)

{
  "mcpServers": {
    "cia-diagnose": { "url": "https://audit.univercityaiconsult.tech/mcp" }
  }
}

Environment Variables

Variable Default Description
CIA_TRANSPORT stdio Transport: stdio, sse, streamable-http
CIA_HTTP_PORT 3792 HTTP port for remote mode
CIA_DB_PATH ~/.cia-diagnose/sessions.db SQLite database path
CIA_N8N_WEBHOOK (empty) n8n webhook URL for lead capture
CIA_TELEGRAM_BOT_TOKEN / CIA_TELEGRAM_CHAT_ID (empty) Direct Telegram lead alerts
CIA_VAULT_LEAD_LOG (empty) Append-only JSONL path for every lead intake
CIA_RATE_FREE 5 Free diagnoses per client IP per day
CIA_LANG es Default language (es/en)

Every completed business_diagnose forwards the lead (company, industry, team_size, revenue_leak_score, top_actions, pain_points, decision_maker_role, the full diagnosis JSON, source, timestamp) to all configured sinks. Missing config never fails the diagnosis — it just logs and continues.

Industry Benchmarks (YAML-driven)

Add an industry = add a YAML in src/cia_diagnose/domain/diagnosis/benchmarks/. No code changes. Current: construction, healthcare, agency, ecommerce, startup, enterprise, restaurant, real_estate, generic.

Brand assets

brand/ holds the official CIA logos (monogram light/dark, gothic wordmark, favicons). Wherever a logo appears it links to https://www.univercityaiconsult.tech/.

License

MIT — © 2026 CIA — Consultoría de Inteligencia Aplicada (David Lopez). See LICENSE.

Recommended Servers

playwright-mcp

playwright-mcp

A Model Context Protocol server that enables LLMs to interact with web pages through structured accessibility snapshots without requiring vision models or screenshots.

Official
Featured
TypeScript
Magic Component Platform (MCP)

Magic Component Platform (MCP)

An AI-powered tool that generates modern UI components from natural language descriptions, integrating with popular IDEs to streamline UI development workflow.

Official
Featured
Local
TypeScript
Audiense Insights MCP Server

Audiense Insights MCP Server

Enables interaction with Audiense Insights accounts via the Model Context Protocol, facilitating the extraction and analysis of marketing insights and audience data including demographics, behavior, and influencer engagement.

Official
Featured
Local
TypeScript
VeyraX MCP

VeyraX MCP

Single MCP tool to connect all your favorite tools: Gmail, Calendar and 40 more.

Official
Featured
Local
graphlit-mcp-server

graphlit-mcp-server

The Model Context Protocol (MCP) Server enables integration between MCP clients and the Graphlit service. Ingest anything from Slack to Gmail to podcast feeds, in addition to web crawling, into a Graphlit project - and then retrieve relevant contents from the MCP client.

Official
Featured
TypeScript
Kagi MCP Server

Kagi MCP Server

An MCP server that integrates Kagi search capabilities with Claude AI, enabling Claude to perform real-time web searches when answering questions that require up-to-date information.

Official
Featured
Python
E2B

E2B

Using MCP to run code via e2b.

Official
Featured
Neon Database

Neon Database

MCP server for interacting with Neon Management API and databases

Official
Featured
Exa Search

Exa Search

A Model Context Protocol (MCP) server lets AI assistants like Claude use the Exa AI Search API for web searches. This setup allows AI models to get real-time web information in a safe and controlled way.

Official
Featured
Qdrant Server

Qdrant Server

This repository is an example of how to create a MCP server for Qdrant, a vector search engine.

Official
Featured