microsoft/flint-chart

What it solves

Flint addresses the difficulty AI agents and developers face when creating polished, professional charts. Instead of requiring the manual tuning of verbose settings for axes, scales, spacing, and layout—which is often too complex for agents to handle reliably—Flint provides a compact, human-editable intermediate language that automatically derives these visual decisions from the data and semantic types.

How it works

Flint acts as a compiler that transforms a simplified chart specification into native code for various visualization libraries. It uses semantic types (over 70 types like Price or Country) and optional visual themes to automatically determine the best layout and presentation.

The project consists of two primary components:

• flint-chart: A library that compiles Flint specs into outputs for Vega-Lite, ECharts, Chart.js, Plotly, or native Excel charts.
• flint-chart-mcp: A Model Context Protocol (MCP) server that allows AI agents to create, validate, and render these charts directly within a chat or coding environment.

Who it’s for

• AI Agent Developers: Who want to give their agents the ability to generate high-quality, brand-consistent visualizations without complex prompt engineering for every chart library.
• Frontend Developers: Who need a unified way to generate charts across multiple different rendering backends.
• Data Analysts: Who want to create polished charts using a simplified specification language.

Highlights

• Multi-Backend Support: Single input compiles to Vega-Lite, ECharts, Chart.js, Plotly, and Excel.
• Semantic Awareness: Uses 70+ semantic types to automatically adapt chart design to the nature of the data.
• Formal Visual Themes: Includes built-in presets (e.g., New York Times, Economist, McKinsey) to ensure brand consistency.
• Agent-Ready: Integrated MCP server for seamless agent-driven chart authoring and interactive previews.