🚢 TitanicAIAnalysis: Análisis de Datos con Claude y MCP

What is 🚢 TitanicAIAnalysis: Análisis de Datos con Claude y MCP?

It is an educational MCP server that exposes the famous Titanic dataset (from Kaggle) as structured resources and tools for analysis with Claude. Designed for developers and data analysts who want to learn how to integrate structured data with large language models using the Model Context Protocol.

How to use 🚢 TitanicAIAnalysis: Análisis de Datos con Claude y MCP?

Install Python 3.10+, clone the repository, create a virtual environment with uv, install dependencies (mcp[cli] and pandas), and ensure Titanic.csv is present. Run the server directly with python main.py, or configure it in Claude Desktop using the provided start scripts (inicio.sh/inicio.bat) or via mcp install main.py. Once connected, Claude can access resources and tools via the MCP interface.

Key features of 🚢 TitanicAIAnalysis: Análisis de Datos con Claude y MCP

• Two resources: full passenger dataset and pre-calculated survival statistics.
• One tool to search for passengers by name.
• Dynamic data loading on each request.
• Structured data exposure for LLM analysis.
• Easy integration with Claude Desktop.
• Educational example of MCP server implementation.

Use cases of 🚢 TitanicAIAnalysis: Análisis de Datos con Claude y MCP

• Ask Claude for overall survival statistics from the Titanic.
• Compare survival rates between passenger classes.
• Search for specific passengers (e.g., surname "Astor").
• Analyze correlations like gender, class, and fare vs survival.
• Generate simple textual analysis without manually uploading files.

FAQ from 🚢 TitanicAIAnalysis: Análisis de Datos con Claude y MCP

What makes this different from just sending a CSV file to Claude?

The server exposes data as structured MCP resources, allowing Claude to access pre-processed statistics and use a search tool dynamically—without manual file uploads. Data is always fresh from the server.

What are the runtime requirements?

Python 3.10 or higher, the mcp[cli] and pandas packages, and Claude Desktop (for integration). The Titanic.csv dataset is included in the repository.

Where does the data come from?

The dataset is the classic Titanic passenger list from Kaggle, provided as Titanic.csv in the repository. No external API is used.

Are there any limitations?

Large datasets may exceed Claude’s context window. Resources are read on each request (no caching). Advanced visualizations are not built-in; you would need to add custom tools.

What transport or authentication does the server use?

It uses the standard MCP protocol over stdio. No authentication is implemented—designed for local, trusted environments.