
WaniKani
io.github.jbeshir由 GitHub 帳號 jbeshir 發布v1.0.1更新於 Oct 29, 2026
Read WaniKani SRS data: profile, reviews, assignments, subjects, and level progressions.
概覽
唯讀的 MCP 伺服器,讓助理查詢 WaniKani 使用者的日語間隔重複學習資料:個人檔案、課程與複習排程、作業、條目與等級進度。
- 可以做什麼
- 讀取使用者檔案、等級、訂閱與休假狀態
- 查看依小時分組的可用課程與複習
- 依條目類型、等級、SRS 階段與可用性篩選 SRS 作業
- 查詢部首、漢字與詞彙的含義、讀音與記憶法
- 查看各條目的複習正確率統計,包括正確/錯誤次數與連續紀錄
- 取得附時間戳的等級晉升歷史
- 適用情境
- 適合學習者請助理彙整 WaniKani 的學習負擔、說明條目為何到期、查詢漢字或詞彙細節,或在不開啟網站的情況下追蹤正確率與等級進度。
- 執行需求
- 以本機程序(stdio)在桌面執行;需要 WANIKANI_API_KEY 環境變數中的 WaniKani API 權杖,並需能連線至 WaniKani API 的網路。從原始碼建置需要 Go;亦提供預先編譯的執行檔與 mcpb 套件。
安裝
在 SourceWeft 中
- 開啟 儀表板中的 WaniKani,將其新增到工作區。
- 為需要使用其工具的對話啟用該服務。
Desktop only,透過 STDIO。 STDIO 服務會啟動本機處理程序,因此需要 SourceWeft 桌面主機。
其他 MCP 客戶端
參照 儲存庫 中的啟動說明。
README
wanikani-mcp
An MCP server that provides read-only access to a user's WaniKani data. WaniKani is a spaced repetition system (SRS) for learning Japanese kanji and vocabulary. This server exposes six tools that let Claude and other MCP clients query a user's profile, lesson/review schedule, SRS assignments, subject details, review statistics, and level progression history.
Key Concepts
- SRS (Spaced Repetition System) -- WaniKani's core learning mechanism. Items progress through stages 0 (initiate) to 9 (burned) based on correct and incorrect reviews. Higher stages mean longer intervals between reviews.
- Subjects -- The items a user learns. There are four types: radicals (building blocks), kanji (characters), vocabulary (multi-kanji words), and kana vocabulary (kana-only words). Each subject has meanings, readings, and mnemonics.
- Assignments -- Track a user's SRS progress on each subject: current stage, when it was unlocked, started, passed, and burned.
- Lessons -- The introduction of new subjects. A subject becomes available as a lesson when its prerequisites are met.
- Reviews -- Periodic quizzes that test recall. Correct answers advance the SRS stage; incorrect answers regress it.
- Levels -- A user's overall progress through WaniKani (1--60). Advancing to a new level unlocks new subjects.
Architecture
The server has three layers: an entry point that reads configuration and wires dependencies, an MCP server layer that registers tools and converts parameters, and an HTTP client that communicates with the WaniKani API.
The server uses mcp-go for MCP protocol handling and stdio transport. The WaniKani API client uses Go generics (Resource[T] and Collection[T]) to provide typed access to all endpoints with automatic pagination.
Data Flow
Collection endpoints are automatically paginated. Results are capped at a configurable limit (default 500, max 10,000) and include a "Showing N of M" header so the client knows whether results were truncated.
Getting Started
Requirements
- Go 1.27.3+ (to build from source)
- WaniKani API token from wanikani.com/settings/personal_access_tokens
Installation
Pre-built binaries are available from Releases.
Install from source:
Or build from the repo root:
Configuration
Claude Desktop
Claude Code
Tools
All tools are read-only. Collection tools accept a limit parameter (default: 500, max: 10,000). Filter parameters that accept multiple values use comma-separated strings (e.g. "radical,kanji").
來源:wanikani/README.md,提交 27dc1b2
工具
0版本歷史
2- v1.0.1最新Oct 29, 2026
- v1.0.0Sep 16, 2026
評分
已滿足適用檢查項的 50%
該評分衡量此收錄條目的完整程度,依據註冊中心條目、端點與公開來源計算得出。它不是對伺服器本身的評價,也不表示伺服器安全或目前可用:可連通性是帶日期的檢查結果,評分也不包含安裝量或使用者評價。
- 近 30 天內可連通25 分
請保持端點能夠回應 MCP 握手;SourceWeft 每 30 天探測一次。
- 已宣告認證方式10 分
請在註冊中心條目中宣告憑證請求標頭,或保持端點能夠回應探測,以便判定其認證方式。
- 已列出工具15 分
請讓伺服器在無憑證時回應 tools/list,或發布可在沙箱中無憑證啟動的軟體包。
- 描述必需15 分
- 圖示10 分
- 文件15 分
- 授權條款10 分