Jotty MCP Server
Provides programmatic access to notes, checklists, and Kanban task boards via the Model Context Protocol, with multi-user support through per-call API keys.
README
Jotty MCP Server
Model Context Protocol (MCP) server for Jotty — provides programmatic access to notes, checklists, tasks (Kanban boards), and categories.
Features
- Notes: Create, read, update, delete markdown notes
- Checklists: Full checklist management with nested items, priorities, timestamps
- Tasks/Kanban: Kanban-style task boards with custom columns and item status management
- Search: Full-text search across notes and checklists
- Multi-user: API key passed per tool call (not env var) — different users can share the same MCP instance
Setup
Local
pip install -r requirements.txt
Stdio mode (default — for local editors/tools like Claude Desktop, Cursor):
JOTTY_BASE_URL=https://jotty.showaib.com python -m src.main
HTTP mode (for remote MCP clients):
JOTTY_BASE_URL=https://jotty.showaib.com PORT=3008 python -m src.main --transport http
Docker
# Stdio mode (default — local CLI)
docker compose up --build -d jotty-mcp
# HTTP mode (remote clients)
MCP_TRANSPORT=http docker compose up --build -d jotty-mcp
Server is available at http://localhost:3008/mcp when using HTTP transport. Connect with any MCP HTTP client.
Configuration
| Variable | Default | Description |
|---|---|---|
JOTTY_BASE_URL |
http://localhost:3000 |
URL of your Jotty instance |
MCP_TRANSPORT |
stdio |
Transport mode (stdio or http) |
PORT |
3008 |
HTTP server port (docker-compose uses host networking) |
Note: The API key is not an environment variable — it is passed per-tool call to support multiple users.
MCP Tools
Notes
| Tool | Description |
|---|---|
list_notes(api_key, category?, q?) |
List notes with optional category filter and search |
get_note(api_key, note_id) |
Get a single note by ID |
create_note(api_key, title, content?, category?) |
Create a new note (content supports Markdown) |
update_note(api_key, note_id, title?, content?, category?) |
Update a note (provide only fields to change) |
delete_note(api_key, note_id) |
Delete a note by ID |
Checklists
| Tool | Description |
|---|---|
list_checklists(api_key, category?, checklist_type?, q?) |
List checklists with optional filters (checklist_type: simple/task) |
get_checklist(api_key, list_id) |
Get a single checklist by ID |
create_checklist(api_key, title, category?, checklist_type?) |
Create a new checklist (use checklist_type='task' for Kanban) |
delete_checklist(api_key, list_id) |
Delete a checklist by ID |
Checklist Items
| Tool | Description |
|---|---|
add_checklist_item(api_key, list_id, text, parent_index?, status?) |
Add item to checklist (use parent_index for nesting) |
update_checklist_item(api_key, list_id, item_index, text?, description?, priority?, score?, start_date?, target_date?, estimated_time?) |
Update item fields |
delete_checklist_item(api_key, list_id, item_index) |
Delete an item by index path (e.g., "0", "0.1") |
check_checklist_item(api_key, list_id, item_index) |
Mark item as completed |
uncheck_checklist_item(api_key, list_id, item_index) |
Mark item as incomplete |
reorder_checklist_items(api_key, list_id, active_item_id, over_item_id, position?, is_drop_into?) |
Reorder items (before/after/nest) |
Tasks (Kanban)
| Tool | Description |
|---|---|
list_tasks(api_key, category?, status?, q?) |
List Kanban task boards |
get_task(api_key, task_id) |
Get a single Kanban board by ID |
create_task(api_key, title, category?, statuses?) |
Create a Kanban board (statuses: optional JSON string of columns) |
update_task(api_key, task_id, title?, category?) |
Update board metadata (title or category only) |
delete_task(api_key, task_id) |
Delete a Kanban board by ID |
Kanban Columns (Statuses)
| Tool | Description |
|---|---|
get_task_statuses(api_key, task_id) |
Get all Kanban columns for a board |
add_task_status(api_key, task_id, status_id, label, color?, order?) |
Add a new Kanban column |
update_task_status(api_key, task_id, status_id, label?, color?, order?) |
Update a column's properties |
delete_task_status(api_key, task_id, status_id) |
Delete a column (items move to first available) |
Task Items
| Tool | Description |
|---|---|
create_task_item(api_key, task_id, text, status?, parent_index?) |
Add item to a Kanban board |
get_task_item(api_key, task_id, item_index) |
Get an item by index |
update_task_item_status(api_key, task_id, item_index, status) |
Move item to different column |
delete_task_item(api_key, task_id, item_index) |
Delete an item by index |
General
| Tool | Description |
|---|---|
search_jotty(api_key, q, search_type?) |
Full-text search across notes and checklists (search_type: note/checklist) |
get_categories(api_key) |
Get all categories organized by type |
get_summary(api_key, username?) |
Get user statistics (note/checklist counts, completion rates) |
Example Usage
# List all notes for a user
mcp.list_notes(
api_key="user1-api-key-here",
category="Work"
)
# Create a note
mcp.create_note(
api_key="user1-api-key-here",
title="Meeting Notes",
content="# Weekly Sync\n- Discussed roadmap\n- Action items...",
category="Work"
)
# Add item to checklist
mcp.add_checklist_item(
api_key="user2-api-key-here",
list_id="<checklist-uuid>",
text="Review PRs",
status="in_progress"
)
# Move task between Kanban columns
mcp.update_task_item_status(
api_key="user1-api-key-here",
task_id="<task-uuid>",
item_index="0",
status="in_progress"
)
License
MIT
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。