Analytical MCP Server

What is Analytical MCP Server?

A Model Context Protocol server exposing 12 analytical tools for Claude, including descriptive statistics, regression (linear/polynomial/logistic/multivariate), hypothesis testing, data preprocessing, data visualization, multi-criteria decision analysis, ML model evaluation, logical argument/fallacy analysis, and Exa-backed research verification. It runs over stdio; the analytical core needs no API key, but research features require an EXA_API_KEY.

How to use Analytical MCP Server?

Install Node.js >=20.0.0, clone the repository, run npm install and npm run build, then configure Claude Desktop to launch the server with the EXA_API_KEY environment variable. Alternatively, use Docker with the docker run command and an .env file. Add the server to Claude Desktop’s mcpServers configuration with the appropriate command and args.

Key features of Analytical MCP Server

• 12 analytical tools for statistical, ML, and logical analysis
• Local analytical processing without an API key
• Optional Exa-backed research verification and perspective generation
• Built-in Prometheus-style metrics server (opt-in)
• Audit logging for every tool call (stderr)
• Layered, namespace-aware caching for research results
• Real statistical hypothesis tests with exact p-value computation

Use cases of Analytical MCP Server

• Analyze datasets with descriptive statistics, regression, and hypothesis tests
• Make multi-criteria decisions using weighted score matrices
• Detect logical fallacies and evaluate argument structure in text
• Verify research claims via multi-source Exa search
• Evaluate ML model predictions with classification or regression metrics

FAQ from Analytical MCP Server

Does the Analytical MCP Server require an API key for all tools?

No. The analytical core (statistics, regression, hypothesis testing, etc.) works without any API key. Only the verify_research and perspective_shifter tools require an EXA_API_KEY and will error at call time if it is missing.

What happens if I call a research tool without EXA_API_KEY?

The tool will error at call time. The tools are registered unconditionally, but they fail gracefully if the key is absent.

How do I enable the metrics server?

Set METRICS_ENABLED=true as an environment variable. The metrics server exposes Prometheus-style metrics, JSON metrics, and a health endpoint on port 9090 by default (configurable via METRICS_PORT and METRICS_HOST).

What are the prerequisites to run the server?

Node.js >=20.0.0 is required. For Docker, any recent Docker installation works. The EXA_API_KEY environment variable is only needed for research features.

Is the decision_analysis tool different from older documentation?

Yes. The tool now requires a scores matrix (options × criteria, values 0–10) in addition to options and criteria. weights is optional and defaults to equal weighting.