Architecture

What this is: the mental model for Publisher and the Malloy stack it sits on. Read it when you want to know which repo owns what, or how a query flows from a .malloy file to a rendered chart.

Publisher is one component of the Malloy stack. From the bottom up:

Malloy

The core compiler and query execution engine. Malloy compiles .malloy files into SQL, executes queries against databases, and returns structured Result objects. Malloy is a pure JavaScript/TypeScript library with no UI or serving capabilities — it's the foundation everything else builds on.

Repository: github.com/malloydata/malloy

Malloy Render

A visualization library that transforms Malloy Result objects into interactive tables, charts, and dashboards. When Malloy executes a query, the result includes both data and rendering hints — tags like # bar_chart or # line_chart that indicate how the data should be displayed. Malloy Render interprets these tags and produces the appropriate visualization.

Built with: SolidJS and Vega/Vega-Lite. Available as both a JavaScript API (MalloyRenderer) and a <malloy-render> web component.

Repository: github.com/malloydata/malloy/packages/malloy-render

Publisher

The analytics engine for Malloy. Publisher serves Malloy models over the network — to agents, applications, and BI tools — and provides a professional UI for data exploration.

• Server: REST API for listing content, managing database connections, compiling models, and executing queries. Also provides an MCP API for AI-agent integration, including the agent retrieval tools and the agent skills as MCP prompts. Supports runtime givens for model-driven, server-side query parameters.
• Console: Web interface for browsing Malloy content, exploring models with a no-code query builder, and viewing results.

Publisher SDK (internal)

A React component library — API communication, query execution, Malloy Render integration, and pre-built pages for browsing environments, packages, models, and notebooks. The Console is built entirely from it, so it's best understood as the Console's internal toolkit rather than a supported external integration path. See embedded-data-apps.md for the advanced/internal notes. To surface analytics, prefer the Publisher Console, an HTML data app, or the REST/MCP APIs.

Credible (hosted)

Credible creates and maintains Publisher, and runs it as the core of the AI Analytics Engine — the hosted, governed service built around Publisher. Everything described above is the same there: the same Malloy, the same package format, the same REST and MCP surfaces. What Credible adds is the engine around them: managed materialization and indexing in storage it brings along, one gateway that enforces access on every query, a concept index that gives an agent just the slice of a model its question needs, and every surface — agents, dashboards and workspaces, data apps and APIs — served from one model. A package published to a local Publisher and one published to Credible are the same artifact; the publisher connection type is how a local authoring loop talks to a hosted environment.

Read more: credibledata.com/malloy · Inside the AI Analytics Engine

Packages in this repo

Package	Description
packages/server	Express.js backend providing REST API (port 4000) and MCP API (port 4040). Loads Malloy packages, compiles queries, executes against databases.
packages/sdk	React component library the Console is built from (internal — see above).
packages/app	Reference implementation and production-ready data exploration tool built with the SDK.
packages/python-client	Auto-generated Python SDK for the REST API.