Artefact Revenue

You can run the Artefact Revenue MCP Server to give AI agents access to revenue intelligence tools like RFM analysis, ICP qualification, and pipeline health scoring. This server is designed to work with MCP clients and can operate with demo data or live HubSpot integration, enabling practical, automated insights for B2B revenue workflows.

How to use

To use the Artefact Revenue MCP Server, connect it to your MCP client (such as Claude or Cursor) and choose one of the available actions: run an RFM analysis, qualify a prospect with the 14.5-point ICP model, or score your pipeline health. You can start with demo data to explore built-in capabilities, then switch to live HubSpot data by providing a HubSpot API key.

How to install

Prerequisites: Python installed on your system, and optionally Node tools if you plan to use MCP clients that rely on npm/npx. You will also need a HubSpot API key if you want to connect to live data.

# Install the MCP server package
pip install artefact-mcp

# If you prefer the Smithery workflow, install the MCP client via Smithery
npx @smithery/cli install artefact-revenue-intelligence

Additional setup notes

If you want to use the live HubSpot data, you will need a HubSpot private app token. The server also supports a demo data mode that uses source=sample so you can experiment without connecting to HubSpot.

Configuration and usage notes

Configure your MCP client to load the Artefact Revenue MCP Server using a standard MCP command flow. The server can be run through an MCP runtime such as uvx. You can keep your HubSpot API key in environment variables for secure access.

Examples of common commands to start using the MCP server

# Start the MCP server via uvx with the artefact-mcp
uvx artefact-mcp

# Or via a practical client setup in your environment (example for the Claude plugin)
claude mcp add artefact-revenue -- uvx artefact-mcp

Available tools

run_rfm

Performs Recency, Frequency, and Monetary value analysis to classify customers into 11 segments and extract ICP patterns from top performers.

qualify_prospect

Applies a 14.5-point model across Firmographic, Behavioral, and Strategic Fit to classify prospects into Ideal/Strong/Moderate/Poor with an engagement strategy.

score_pipeline

Evaluates open deals to produce a 0-100 health score by analyzing velocity, bottlenecks, and stage transitions.