truefold
MCP server

Your metrics, in the AI tool you already use.

Truefold exposes your semantic layer over the Model Context Protocol. Point Claude or any MCP-aware client at it and it can describe your data model, resolve a metric, list what a dimension filters by, and run a checked query — against the same definitions the app uses. You are not locked into our chat window.

01 · Bring your own AI

Your metrics, inside the AI tool you already use.

Truefold exposes your semantic layer over the Model Context Protocol. Point Claude — or any MCP-aware client — at it and it can describe your data model, resolve a metric, list what a dimension can be filtered by, and run a query. Same definitions as the built-in chat, so the two can't give different answers.

  • describe_data_model: what tables exist, what they mean, how they relate
  • get_metric_value: resolve a named metric, optionally broken down by a dimension
  • list_dimension_values: what a filter can actually be set to
  • run_query / validate_sql: for teams who want to write their own, checked against the sanctioned joins
  • Per-user tokens: each token maps to one person in one organisation, revocable, with an expiry
  • Hosted: no infrastructure to run on your side
In practice

Your team lead asks Claude every Monday: any channel whose contribution margin dropped more than five points week on week? It resolves against the same definitions your dashboards use, so the answer matches the one in the app — and the coverage caveat comes along with it.

Architecture
Your AI client
ClaudeAny MCP-aware clientYour own agents
MCP · bearer token
Truefold MCP server
Nine tools over your layer · per-user token, scoped to one organisation, revocable
Your reviewed semantic layer
40 metrics · 14 dimensions · checked joins
The same definitions chat, dashboards and Signals resolve against.
Your sources
ShopifyMeta AdsGoogle AdsKlaviyoor your own BigQuery
  • Read-only on every source.

    Truefold pulls; it never writes. Nothing in your store, your ad accounts or your flows changes because you connected us.

  • Isolated, and minimised on the way in.

    One BigQuery dataset per organisation, used by nobody else. Customer names, email addresses, phone numbers and shipping addresses are dropped before anything is written.

  • The definitions are yours to adjust.

    Rename metrics to your team's language, correct anything that doesn't match how you run the business. Version-controlled, with a name against every change. We don't train AI on your business data.

See it against your own numbers.

Beta access is going out to Shopify brands a few at a time. Tell us what you're running and we'll walk you through what it finds.