MCP Server

Hey There,

I've spent a lot of the last year watching teams bolt AI onto their go-to-market motion and get... nothing useful back. The AI could read the CRM, sure. It just couldn't think about it. Ask "is my pipeline healthy?" and you'd get a polite summary of rows, not an answer.

That gap is what I ended up building Artefact MCP to close.

It's an MCP server that exposes your GTM strategy as something an AI agent can query and act on, thirteen tools across two tiers. It runs in Claude Desktop, Claude Code, Cursor, Windsurf, ChatGPT Desktop, Gemini CLI, and Copilot.

The part I'm most proud of: you can get value before connecting anything. Five tools work with zero HubSpot connection at all, an AI-infrastructure archetype selector, a 34-module GTM playbook catalog, a deployment-mode diagnostic, a CLAUDE.md generator, and a module milestone validator. Four more diagnostics (AI maturity score 0–100, pipeline hygiene audit, lifecycle-drift detection, property-governance audit) run on sample data or your read-only HubSpot key. Nine tools, free for life.

The four Pro tools are where the live revenue intelligence lives: ICP triangulation (firmographics + behavioral signals + growth indicators, not just company size), an 11-segment RFM engine, pipeline health scoring, and a config validator.

How it's different from the official HubSpot MCP: HubSpot's MCP is built for CRM operations, read/write on objects. Artefact isn't trying to compete with that. It's the strategy layer on top: methodology, scoring, decisions. They're complementary. HubSpot for execution, Artefact for the thinking.

Getting started is one line:

pip install artefact-mcp

Then claude mcp add artefact-mcp, and you ask in plain language. "score my HubSpot's AI maturity" or "audit my pipeline hygiene."

Where I'm at: it's published on the MCP registry and live on PyPI. Free tier is genuinely free, Pro is $149/mo, Enterprise $499/mo. It's already being used for real GTM analysis, not just a demo.

I'd love feedback from this crowd specifically — especially anyone running an AI-native sales or RevOps motion. What would make a tool like this indispensable vs. nice-to-have? And does the "free until you connect live data" model resonate, or would you rather just pay upfront for the full thing?

Install: https://pypi.org/project/artefact-mcp/

Code: https://github.com/alexboissAV/artefact-mcp-server

Happy to answer anything in the comments.