Rapiber Ticket Manager
A local-first Kanban system with MCP tools for AI agents to manage tickets, features, and tasks, paired with a React UI for human users.
README
Rapiber Ticket Manager (MCP Server & Kanban UI)
Ticket Manager is a robust, local-first Kanban system designed specifically to be controlled by AI Agents via the Model Context Protocol (MCP), while also providing a beautiful React-based Kanban UI for human engineers.
Quick Start
1. Prerequisites
- Python 3.10+
- Node.js 18+
2. Setup the Project
Clone the repository and set up the environments:
# Setup Python Backend (MCP + FastAPI)
python -m venv venv
.\venv\Scripts\activate
pip install -r requirements.txt
# Setup React UI
cd ui
npm install
npm run build
cd ..
3. Run the Server
The Python backend serves both the API and the React UI bundle.
# Run the FastAPI server
.\venv\Scripts\python -m uvicorn src.api:app --host 127.0.0.1 --port 8000
Navigate to http://127.0.0.1:8000 in your browser to view the Kanban board.
Commands
| Command | Description |
|---|---|
npm run dev |
Start the Vite development server for the UI with HMR |
npm run build |
Compile the React UI for production |
python -m uvicorn src.api:app --reload |
Run the FastAPI backend with auto-reload |
Architecture
This project is built using a modern, lightweight, and highly portable stack:
- Database:
SQLite3- Chosen for absolute portability. No external Docker containers or DB servers are required. - Agent Protocol:
FastMCP- Exposes the database to AI agents natively via their IDEs. - REST API:
FastAPI- A highly secure, rate-limited, and strict CORS-enforced API serving the human UI. - Frontend UI:
React + Vite + TailwindCSS- A responsive, beautiful Kanban board for manual human intervention and review.
Available MCP Tools
When mounted in an AI Agent, this server exposes the following tools to the LLM context:
| Tool Name | Description |
|---|---|
create_feature |
Create a new parent feature/epic (Requires Manager role) |
get_feature |
Retrieve details of a specific feature by ID |
list_features |
List all available features |
create_subfeature |
Create a logical grouping inside a feature |
get_subfeature |
Retrieve details of a subfeature |
create_ticket |
Create a new ticket (Requires Manager role) |
get_ticket |
Retrieve full context, tasks, and historical notes for a specific ticket ID |
list_tickets |
Powerful search tool to filter tickets by status, assignee, priority, type, parent, or text |
update_ticket_status |
Move a ticket across the Kanban board (Enforces role-based state transitions) |
add_ticket_task |
Add a sub-task checklist item to a ticket |
check_ticket_task |
Mark a sub-task as completed (Required before moving to IN_REVIEW) |
add_ticket_note |
Append a note/comment to the ticket's activity log |
assign_ticket |
Assign a ticket to an engineer |
AI Agent Instructions (mcp.json)
To use this server with an AI Agent (like Claude Desktop or Antigravity), you need to mount it in your mcp.json configuration file.
Add the following entry to your mcpServers block. Be sure to replace the cwd path with the actual absolute path to the ticket-mcp-server directory on your machine:
{
"mcpServers": {
"ticket-manager": {
"command": "python",
"args": [
"src/server.py"
],
"cwd": "/absolute/path/to/mcp-kanban",
"env": {}
}
}
}
Once mounted, the agent will have access to tools like list_tickets, get_ticket, add_ticket_note, and update_ticket_status.
Contributing
When modifying the backend, ensure that all database interactions run through the context manager in src/db.py to prevent connection leaks.
推荐服务器
Baidu Map
百度地图核心API现已全面兼容MCP协议,是国内首家兼容MCP协议的地图服务商。
Playwright MCP Server
一个模型上下文协议服务器,它使大型语言模型能够通过结构化的可访问性快照与网页进行交互,而无需视觉模型或屏幕截图。
Magic Component Platform (MCP)
一个由人工智能驱动的工具,可以从自然语言描述生成现代化的用户界面组件,并与流行的集成开发环境(IDE)集成,从而简化用户界面开发流程。
Audiense Insights MCP Server
通过模型上下文协议启用与 Audiense Insights 账户的交互,从而促进营销洞察和受众数据的提取和分析,包括人口统计信息、行为和影响者互动。
VeyraX
一个单一的 MCP 工具,连接你所有喜爱的工具:Gmail、日历以及其他 40 多个工具。
graphlit-mcp-server
模型上下文协议 (MCP) 服务器实现了 MCP 客户端与 Graphlit 服务之间的集成。 除了网络爬取之外,还可以将任何内容(从 Slack 到 Gmail 再到播客订阅源)导入到 Graphlit 项目中,然后从 MCP 客户端检索相关内容。
Kagi MCP Server
一个 MCP 服务器,集成了 Kagi 搜索功能和 Claude AI,使 Claude 能够在回答需要最新信息的问题时执行实时网络搜索。
e2b-mcp-server
使用 MCP 通过 e2b 运行代码。
Neon MCP Server
用于与 Neon 管理 API 和数据库交互的 MCP 服务器
Exa MCP Server
模型上下文协议(MCP)服务器允许像 Claude 这样的 AI 助手使用 Exa AI 搜索 API 进行网络搜索。这种设置允许 AI 模型以安全和受控的方式获取实时的网络信息。