Spec Driven Development
MCP Server that facilitates spec-driven development workflows, not just Vibe Coding.
作者提供的说明(英文原文)
Read Bunpro SRS data: profile, study queue, reviews, JLPT progress, and grammar lookups.
这三项永远不会合并成一个分数:意见、测量和人气是不同的东西。
我们还没有为它写通俗的介绍。下面是我们从注册中心读取到的事实。
选择你使用的 AI 应用,然后复制设置。
BUNPRO_EMAIL, BUNPRO_PASSWORD标为机密的值是你账户的钥匙。千万不要分享给别人,也不要粘贴到聊天里。
你用的是哪个应用?
在上方选择你的应用。我们会记住它,下次按钮会显示“安装到 …”。
链接指向作者自己的文件;我们不托管也不重新打包它。打开文件后,Claude Desktop 会进行安装。
在终端中运行这条命令
运行前请先填写空着的值。
浏览器会先询问是否打开应用,然后应用会显示设置并请你确认。密钥的值会留空,由你自己填写。
浏览器会先询问是否打开应用,然后应用会显示设置并请你确认。密钥的值会留空,由你自己填写。
这个服务器在你的电脑上运行,而 ChatGPT 只能连接互联网上的服务器。请选择其他应用。
你的选择通过 Cookie 保存在这台设备上,无需账户。
这些设置是根据作者在注册中心的条目生成的。如果 README 的写法不同,请以 README 为准。
它在你的电脑上运行;安装后,工具列表会显示在你的 AI 应用中。
我们检查的是任何人都能看到的信息;我们没有阅读代码。不适用的检查项不计入总数。 · 检查于 2026年11月7日
注册名称已通过 GitHub 账户 jbeshir 验证,该账户拥有此仓库
我们还没有从 GitHub 读取这个仓库
我们还没有从 GitHub 读取这个仓库
它在你的电脑上运行,没有互联网端点
在你的电脑上运行,改用本地设置
它在你的电脑上运行;我们只能读取互联网端点的工具列表
与注册中心里这个分类中所有活跃的服务器比较。只有当发布者拥有该仓库时,星标数才计入。
它在 1 项适用的基础检查中通过了 1 项。我们没有阅读代码。在添加密钥或令牌之前,请先看看这个服务器要求哪些权限,并使用只具备所需权限的密钥。
你需要填写:BUNPRO_EMAIL, BUNPRO_PASSWORD
它在你的电脑上运行。你的 AI 应用会在需要时启动它,数据不会经过作者的服务器(除非它自己调用在线服务)。
GitHub 仓库最后更新于–,最新版本(1.0.1)于12天前进入注册中心。
支持 MCP 的应用,例如 Claude Desktop、Claude Code、Cursor 和 VS Code。你的电脑上还需要有 MCPB。
同一分类中有:Spec Driven Development, Remote Desktop Commander, Signoff
同一分类中其他作者的服务器,按星标数排序。
如何看卡片:★ GitHub 星标i三大支柱:评价 · 检查 · 使用情况i更新 = 最新版本发布到注册中心的时间i
MCP Server that facilitates spec-driven development workflows, not just Vibe Coding.
Remote Desktop Commander: use files, terminal and processes on your own computers from any AI app.
Five gates that make an agent's 'done' provable: run them, read the verdict, audit the record.
按原文从作者的仓库复制。PopNx MCP 没有审查或背书其内容。HTML、图片和脚本已被移除。链接会离开本站。
An MCP server that provides read-only access to Bunpro, a Japanese grammar and vocabulary SRS (spaced repetition system) platform. It exposes 12 tools for querying a user's profile, study statistics, review forecasts, JLPT progress, SRS stage details, and individual grammar point or vocabulary data. Authentication is handled automatically via Bunpro's Rails Devise login flow using your email and password.
The server has four layers: an entry point that reads configuration and obtains a token, an auth module that performs the Devise login flow when needed, an HTTP client that calls the Bunpro frontend API with Bearer token auth, and the MCP server that registers tools and formats responses.
graph TD
MCP["MCP Client (Claude, etc.)"] SRV["MCP Server 12 read-only tools"] CLI["HTTP Client Bearer token auth"] API["Bunpro Frontend API api.bunpro.jp"]
MCP -->|"stdio JSON-RPC"| SRV SRV --> CLI CLI -->|"GET /api/frontend/*"| API
subgraph "Authentication (startup)" direction LR LOGIN["Auth Module Devise login flow"] SITE["bunpro.jp Rails login page"] LOGIN -->|"CSRF + POST credentials"| SITE SITE -->|"frontend_api_token cookie"| LOGIN end
LOGIN -.->|"token"| CLI
Authentication happens once at startup. The auth module fetches a CSRF token from the login page, submits form-encoded credentials, and extracts the frontend_api_token cookie. The token has approximately a 2-day lifetime — long enough for any single MCP session.
The sequence diagram below shows both the startup token acquisition and a runtime tool call. All API requests are GET requests to api.bunpro.jp/api/frontend/* with the token sent as a Bearer authorization header.
sequenceDiagram
participant Client as MCP Client participant Server as MCP Server participant Auth as Auth Module participant Site as bunpro.jp participant API as api.bunpro.jp
Note over Client,API: Startup: Token Acquisition Server->>Auth: Login(email, password) Auth->>Site: GET /login Site-->>Auth: HTML with CSRF token Auth->>Site: POST /users/sign_in (form-encoded) Site-->>Auth: frontend_api_token cookie Auth-->>Server: token
Note over Client,API: Runtime: Tool Calls Client->>Server: get_grammar_point(id) Server->>API: GET /api/frontend/reviewables/grammar_point/{id} Note over Server,API: Authorization: Bearer token API-->>Server: JSON:API response Server-->>Client: Formatted JSON result
Pre-built binaries are available from Releases.
Install from source:
go install github.com/jbeshir/mcp-servers/bunpro/cmd/bunpro-mcp@latest
Or build from the repo root:
make build
| Variable | Required | Description |
|---|---|---|
BUNPRO_EMAIL | Yes | Bunpro account email |
BUNPRO_PASSWORD | Yes | Bunpro account password |
BUNPRO_API_URL | No | API base URL (default: https://api.bunpro.jp) |
BUNPRO_LOGIN_URL | No | Login page URL (default: https://bunpro.jp) |
{
"mcpServers": { "bunpro": { "command": "/path/to/bunpro-mcp", "env": { "BUNPRO_EMAIL": "you@example.com", "BUNPRO_PASSWORD": "your-password" } } } }
claude mcp add bunpro /path/to/bunpro-mcp -e BUNPRO_EMAIL=you@example.com -e BUNPRO_PASSWORD=your-password
All tools are read-only. No tool modifies Bunpro data.
| Tool | Parameters | Description |
|---|---|---|
get_user | -- | User profile: level, XP, streak, subscription status |
get_study_queue | -- | Due grammar and vocabulary review counts |
get_decks | -- | Deck settings: daily goals, progress, batch sizes |
get_stats | -- | Study statistics: days studied, streak, accuracy, badges |
get_jlpt_progress | -- | JLPT N5-N1 progress with SRS stage counts |
get_review_forecast | granularity (optional): "daily" or "hourly" | Upcoming review forecast split by grammar and vocabulary |
get_srs_overview | -- | Aggregate SRS level counts (beginner through master, plus ghost and self-study) |
get_review_activity | -- | Daily review counts for the past ~30 days |
get_grammar_srs_details | level (required): beginner, adept, seasoned, expert, or master; page (optional) | Paginated grammar review items at a specific SRS level |
get_vocab_srs_details | level (required): beginner, adept, seasoned, expert, or master; page (optional) | Paginated vocabulary review items at a specific SRS level |
get_grammar_point | id (required) | Grammar point details: meaning, structure, nuance, study questions |
get_vocab | slug_or_id (required): slug (e.g. "食べる") or numeric ID | Vocabulary details: readings, pitch accent, frequency, JMDict entries |