# MAGI Code Review

> Use this tool when you need to automate code reviews with multiple perspectives, leveraging AI assistants to evaluate code quality without leaving their interface. It solves the problem of manual code review by connecting to a distributed review framework with specialized agents, providing a final verdict based on majority rule. The MAGI Code Review tool accepts code submissions and authentication tokens as input, producing a detailed log and final verdict as output.

Canonical page: https://skillsregistry.net/skills/miki-hoshizaki-magi-code-review  
JSON: https://api.skillsregistry.net/v1/skills/miki-hoshizaki-magi-code-review

## Description

MAGI MCP Server provides a code review orchestration system that connects AI assistants to a distributed review framework with three specialized agents (Melchior, Balthasar, and Casper). The server establishes WebSocket connections to the MAGI Gateway, submits code for evaluation, and aggregates agent decisions into a final verdict based on majority rule. Built with FastMCP and supporting both SSE and WebSocket transports, it features authentication token generation, detailed logging, and containerization via Docker. This implementation is particularly valuable for developers seeking automated, multi-perspective code quality assessment without leaving their AI assistant interface.

## Trust

- **Trust score (0–1):** 0.94
- **Verification tier:** verified
- **Last scanned:** 2026-09-28

## Facts

- **Version:** 1.0.0
- **Skill type:** atomic
- **Execution layer:** mcp-remote
- **Runtime environment:** api
- **Category:** cloud-infra
- **Updated:** 2026-09-28

## Source

- **Source listing:** [PulseMCP](https://www.pulsemcp.com/servers/miki-hoshizaki-magi-code-review)
- **Repository:** <https://github.com/miki-hoshizaki/mcp-magi>

## Use it

Resolve this record through the SkillsRegistry MCP server (no auth, read-only):

```
claude mcp add --transport http --scope user skillsregistry https://api.skillsregistry.net/mcp
```

```json
{
  "jsonrpc": "2.0",
  "id": 1,
  "method": "tools/call",
  "params": {
    "name": "get_skill",
    "arguments": {
      "slug": "miki-hoshizaki-magi-code-review"
    }
  }
}
```

REST: `GET https://api.skillsregistry.net/v1/skills/miki-hoshizaki-magi-code-review` · pull for local use: `GET https://api.skillsregistry.net/v1/skills/miki-hoshizaki-magi-code-review/pull`

---
SkillsRegistry indexes agent skills from public registries and GitHub. Skills we have analysed are scanned with Circle-IR and scored on six dimensions; each listing states its scan coverage. More: https://skillsregistry.net/llms.txt