repowiserepowise

Command-line tool

shinkuan / akagi

支持雀魂、天鳳、麻雀一番街、天月麻將,能夠使用自定義的AI模型實時分析對局並給出建議,內建Mortal AI作為示例。 Supports Majsoul, Tenhou, Riichi City, Amatsuki, with the ability to use custom AI models to analyze games in real time and provide suggestions. Comes with Mortal AI as a built-in example.

★ 1,141stars on GitHubRustApache-2.0

Claude Code, Cursor, Codex or VS Code. No key. You approve one sign-in in your browser.

Repowise maps public repos from their source. Not affiliated with shinkuan. What is Repowise?

github.com/shinkuan/akagiIndexed at 97478d0

What is akagi?

akagi is a real-time mahjong AI assistant for Mahjong Soul, Tenhou and other platforms. It watches a game over a local MITM proxy or a Chromium window that it launches, mirrors the game state, and shows shanten, waits, agari rate, tenpai rate, per-opponent deal-in risk and a recommended discard in a draggable HUD. Its core is a single Rust binary built with Tauri, with a TypeScript frontend. Its code health is 6.0 out of 10 (Fair), ranked #215 of 4,000+ indexed repos. The last commit to akagi was on Aug 13, 2026, and 16 people have contributed.

As of commit 97478d0, indexed Aug 14, 2026. What it is comes from a model reading the README; the rest is measured from the source and git history, not from the maintainer.

For agentsakagi.md

Repo facts: type, license, health, rank, contributors
akagi facts
TypeCommand-line tool
LicenseApache-2.0
Code health6.0 / 10 (Fair)
Health rank#215 of 4,000+ indexed repos
Last commitAug 13, 2026
Contributors16
Bus factor1
Size352 files, 76,232 lines

Explore akagi

Ask about this code in your agent

This connects Repowise to akagi's code. It does not install akagi.

Useful if you want to change or fork akagi: your agent can ask how it is put together and get answers that cite its files.

Ask about akagi

Answers come from akagi's README and source, with the files they cite. Two free answers a day without an account.

Want this for your own codebase? Map your repo free

About this page

Get a page like this for your own codebase

Repowise built this map of shinkuan/akagi from its source and git history. It parses every symbol, draws the dependency graph, scores each file's health from complexity, duplication, coverage and churn, finds hotspots and owners in the history, and writes documentation for every module. Your coding agent reads the same index over MCP, so its answers cite real files.

Free for 2 public repos, with a page like this one. No card. See how Repowise works

Prefer local? pip install repowise && repowise init indexes your own checkout, no account needed. Star Repowise on GitHub

shinkuan/akagi FAQ

What is akagi?

akagi is a real-time mahjong AI assistant for Mahjong Soul, Tenhou and other platforms. It shows shanten, waits, agari rate, tenpai rate, per-opponent deal-in risk and a recommended discard in a draggable HUD.

How does akagi work?

akagi captures game traffic through a local MITM proxy or a Chromium window it launches, and mirrors the game state. A bot, either the built-in neural net or a cloud model, returns a suggestion each turn that appears in the HUD.

What language is akagi written in?

akagi is written mainly in Rust, with a TypeScript frontend built on Tauri. Its core is a single Rust binary.

How is the akagi code organized?

The akagi core is a single Rust binary whose subsystems communicate only through the event bus defined in src/event_bus.rs. The UI lives in the frontend directory and the built-in bot lives in native_bot.

Who is akagi for?

akagi is designed for players who want to understand their performance in mahjong matches in real time and learn from it. The project is intended for educational purposes only.

What license does akagi use?

akagi is licensed under Apache-2.0.

Code health

5.9out of 10Fair

Hotspots average 3.7, the place fixes pay off most.

On the leaderboards:#14 of 24 Rust repos

Key features of akagi

  • A draggable HUD shows shanten, waits, agari rate, tenpai rate, per-opponent deal-in risk and a suggested attack or defence discard.
  • Capture works through a local MITM proxy or a Chromium window that akagi launches, and the Chromium mode needs no certificate setup.
  • A built-in pure-Rust neural net bot is included, and cloud inference can be set up as an optional source of decisions.
  • Each completed match is recorded in a game history with rank pie charts, PT line charts and detailed stats.
  • Sanma (3-player) games are supported with AI analysis, per-mode bot routing and history stats.
  • The interface is available in English, Japanese, Traditional Chinese and Simplified Chinese, and the language can be switched live.

How shinkuan/akagi works

Game traffic enters through the MITM proxy or through frames the Chromium window intercepts over the Chrome DevTools Protocol. src/event_bus.rs defines the channel types that subsystems use, and src/lib.rs wires those buses at boot, so each subsystem holds only its own bus handles. The built-in bot's entry point is native_bot/src/lib.rs, which provides a decision each turn that the frontend/src/main.tsx app shows in the HUD through push events such as bot-response. With AutoPlay enabled, src/autoplay/majsoul/mod.rs carries out the bot's decisions in the game client. Bots installed from GitHub or a ZIP are extracted under mjai_bot, where bot.py is validated before use.

The codebase has 352 files and 4,996 symbols in 8 modules, led by Rust, TypeScript and Luau. Code health is 5.9 out of 10, rated fair.