This server acts as an MCP (Model Context Protocol) gateway, empowering AI agents like Claude with the ability to formulate and solve complex combinatorial optimization problems. It integrates with MiniZinc, a high-level constraint modeling language, enabling agents to tackle challenges such as scheduling, resource allocation, graph colouring, packing, and planning. The tool provides a streamlined API for agents to validate models, inspect their interfaces, and execute solves, supporting various input formats including inline code, DZN strings, and JSON data, across both stdio and HTTP transports.
Key Features
01Exposes MiniZinc constraint solving to AI agents via MCP.
02Validates MiniZinc models for syntax and type errors.
03Inspects model interfaces to identify required inputs and outputs.
04Solves MiniZinc models with support for inline code, include files, DZN, and JSON data.
05Supports flexible deployment with both stdio and HTTP transports (including Docker).
060 GitHub stars
Use Cases
01Enabling AI agents to formulate and solve combinatorial optimization problems.
02Automating complex scheduling, resource allocation, and planning tasks through AI.
03Integrating declarative constraint modeling into AI agent workflows.
