Jij MCP Server
Official# Jij MCP Server
A server that provides tools to support mathematical optimization with JijModeling and quantum computing.
## Overview
The Jij MCP Server contains various tools and utilities designed to assist with the implementation of mathematical optimization models using JijModeling and quantum computing tasks using Qiskit. This platform integrates both domains to provide comprehensive support for computational optimization and quantum programming.
## Installation
1. Clone this repository
2. Install the required dependencies
3. Configure the server as described below
## Features
### JijModeling Support
- Reference information about JijModeling syntax and usage
- Code checking for detection of common issues in JijModeling code
- Model creation assistance with best practices guidance
- Step-by-step workflow for implementing optimization models
### Quantum Computing Support
- Qiskit migration guides from v0.x to v1/v2
- API reference documentation access
- Integration with IBM Quantum Learning Hub tutorials
- Structured workflow for quantum circuit design and execution
## Configuration
The MCP server can be configured in your settings file as follows:
```json
{
"mcpServers": {
"jij": {
"command": "uv",
"args": [
"--directory",
"<YOUR PATH>/jij-mcp-server",
"run",
"jij_mcp/server.py"
]
}
}
}
```
This configuration specifies:
- Server name: `jij`
- Command to run: `uv`
- Arguments:
- `--directory`: Specifies the server location
- Path to your server directory
- `run`: Command to execute the server
- `jij_mcp/server.py`: Server script to run
### Docker version
[](https://insiders.vscode.dev/redirect/mcp/install?name=jij&inputs=%5B%5D&config=%7B%22command%22%3A%22docker%22%2C%22args%22%3A%5B%22run%22%2C%22-i%22%2C%22--rm%22%2C%22--platform%22%2C%22linux%2Famd64%22%2C%22ghcr.io%2Fjij-inc%2Fjij-mcp-server%3Alatest%22%5D%7D)
```json
{
"mcpServers": {
"jij": {
"command": "docker",
"args": [
"run",
"-i",
"--rm",
"--platform",
"linux/amd64",
"ghcr.io/jij-inc/jij-mcp-server:latest"
],
}
}
}
```
## Available Tools
### JijModeling Tools
- `learn_jijmodeling`: Guide to JijModeling syntax and usage
- `jm_check`: Validation tool for JijModeling code
### Qiskit Tools
- `qiskit_v0tov1v2_migration_guide`: Guide for transitioning between Qiskit versions
- `qiskit_v1_api_reference_toc` and `qiskit_v2_api_reference_toc`: API documentation access
- `qiskit_tutorial`: Access to IBM Quantum Learning Hub tutorials
## License
Apache License 2.0
TDQS
Scored across 8 tools
Most tools have distinct purposes, with clear separation between JijModeling and Qiskit domains. However, there is some potential confusion between 'qiskit_v0tov1v2_migration_guide' and 'learn_jijmodeling' as both are educational guides, and 'qiskit_v1_api_reference_toc' and 'qiskit_v2_api_reference_toc' are very similar in function. The descriptions help clarify, but the boundaries could be slightly ambiguous.
The naming is mixed with no consistent pattern. Some tools use snake_case (e.g., 'fetch_as_markdown', 'jm_check'), while others use descriptive phrases with underscores (e.g., 'qiskit_v0tov1v2_migration_guide'). There is no uniform verb_noun structure, and names vary in length and style, making them less predictable.
With 8 tools, the count is reasonable for a server covering JijModeling and Qiskit support. It's slightly on the higher side but manageable, as each tool serves a specific function without obvious redundancy. The scope is well-defined, though it could be streamlined.
For the JijModeling domain, there is a check tool and a learning guide, but no creation or execution tools, leaving gaps. For Qiskit, there are checking, tutorial, migration, and API reference tools, which cover educational and reference needs but lack direct code execution or quantum computing operations. The surface is functional but incomplete for full workflow support.