Kanboard Python MCP Server & Automation Tools

This project provides a fully-featured Model Context Protocol (MCP) server for interacting with a Kanboard instance via its JSON-RPC API. It exposes Kanboard functionality directly to LLMs (like Claude, Antigravity, etc.).

Additionally, it contains a set of automation scripts for managing Kanboard projects (like swimlane management, AgileIndicators updates, and bulk task creation).

Features

• MCP Server (kanboard_mcp_server.py): A native MCP server exposing Kanboard tools.
  • get_project_master_map: Fetches a bird's eye view of a project, including active tasks and all attached Markdown (.md) files.
  • get_my_dashboard: Fetches a user's dashboard (all active projects, assigned tasks, unread notifications).
  • Standard Task Tools: create_task, update_task, move_task_position, get_tasks, etc.
• Robust Error Handling: Automatic fallback methods (e.g., trying getAllProjectFiles -> getProjectFiles) to handle different Kanboard versions or permission settings without crashing.
• Diagnostic CLI Tools: A built-in --doctor flag to verify network, credentials, and API connections before launching the server.
• Swimlane & Column Management: Reorder swimlanes, rename default lanes, and move tasks efficiently via Python scripts.
• AgileIndicators Integration: Apply Priority (P1-P5) and Complexity/Score (C0-C50) to your Kanboard tasks.

Installation

1. Clone this repository.
2. Create a virtual environment: python3 -m venv venv
3. Activate it: source venv/bin/activate
4. Install dependencies: pip install -r requirements.txt
5. Copy .env.example to .env and fill in your Kanboard credentials. (Never commit your .env file to version control!)

Usage

Running the MCP Server

To use the server with an MCP client (like Antigravity or Claude Desktop):

python kanboard_mcp_server.py

Note: MCP servers communicate over stdio, so you won't see terminal output directly unless you run with diagnostic flags.

Diagnostics

Check if your .env configuration and Kanboard API connection are working:

python kanboard_mcp_server.py --doctor

Server Status

python kanboard_mcp_server.py --status

Automation Scripts

Check the scripts/ directory for ready-to-use Python scripts that automate common Kanboard workflows (e.g., bulk reordering, applying agile metrics).

Kanboard MCP Araçları (Toplam: 21 Araç)

Araç Adı	Kategori	Parametreler	Açıklama
get_my_projects	Proje Yönetimi	``	Lists your own projects in Kanboard.
get_project_by_id	Proje Yönetimi	project_id: int	Gets the details of a single project. Returns an error for unauthorized projects.
get_tasks_by_project	Görev İşlemleri	project_id: int	Gets all tasks in the given project ID.
get_task_details	Görev İşlemleri	task_id: int	Gets all details of a specific task (title, description, assignee, category, date, etc.).
get_user_tasks	Görev İşlemleri	``	Gets the overdue and open tasks of the current API user (or token owner).
get_columns	Proje Yönetimi	project_id: int	Lists the board columns in a project. Used to find IDs for moving tasks or creating tasks.
create_task	Görev İşlemleri	project_id: int, title: str, description: str = "", column_id: int = 0, owner_id: int = 0, swimlane_id: int = 0	Creates a new task. You can only create tasks in projects you are authorized for.
update_task	Görev İşlemleri	task_id: int, title: str = "", description: str = "", owner_id: int = 0, category_id: int = 0, score: int = 0, swimlane_id: int = 0	Updates an existing task.
move_task_position	Görev İşlemleri	project_id: int, task_id: int, column_id: int, position: int = 1, swimlane_id: int = 0	Moves a task to a specific column and position.
create_comment	Dosya & Yorumlar	task_id: int, user_id: int, content: str	Adds a comment to a specific task.
get_all_subtasks	Alt Görevler	task_id: int	Gets all subtasks for a given task.
create_subtask	Alt Görevler	task_id: int, title: str, time_estimated: int = 0	Creates a single subtask.
create_bulk_subtasks	Alt Görevler	task_id: int, titles_csv: str	Creates multiple subtasks from a comma-separated string of titles.
get_all_swimlanes	Proje Yönetimi	project_id: int	Lists all swimlanes for a project.
change_swimlane_position	Proje Yönetimi	project_id: int, swimlane_id: int, position: int	Changes the position of a swimlane.
create_milestone	Diğer	project_id: int, title: str, column_id: int, date_str: str	Creates a milestone task (colored yellow) for a specific date.
schedule_waterfall	Proje Yönetimi	task_ids_csv: str, start_date_str: str, hours_per_subtask: int = 4	Schedules a list of tasks sequentially (waterfall) starting from start_date_str.
get_project_master_map	Proje Yönetimi	project_id: int	Aggregates project metadata, active tasks, and project files into a single context map for Agent Swarm.
create_external_task_link	Bağımlılık & Harici Bağlantılar	task_id: int, url: str, dependency: str, type: str = "weblink", title: str = ""	Creates an external link for a task.
create_task_link	Bağımlılık & Harici Bağlantılar	task_id: int, opposite_task_id: int, link_id: int	Creates a link between two tasks.
close_task	Görev İşlemleri	task_id: int	Closes a task (sets is_active to 0).

SDLC Tarihçe Aktarıcısı (CLI)

python scripts/setup_project_tasks.py --project-id 7 --repo-path ~/dediaiproje/dedisoft/kanboard/plugins/ExecutiveDashboard --memory-dir docs/ai_chats/

Not: docs/ai_chats/ klasörü .gitkeep ile korunmaktadır ve içerisindeki yapay zeka chat logları (.md) izole bir şekilde saklanıp git tarafından yok sayılır (veya proje hafızası olarak Kanboard'a yüklenir).