zube-mcp

zube-mcp

MCP server for the Zube.io project management API, exposing boards, cards, epics, tickets, sprints, and workspaces as tools for AI assistants.

Category
访问服务器

README

zube-mcp

MCP server for the Zube.io project management API. Exposes Zube boards, cards, epics, tickets, sprints, and workspaces as tools that AI assistants can call.

Setup

Prerequisites

  • Python 3.10+
  • uv:
    • macOS: brew install uv
    • Windows: powershell -ExecutionPolicy ByPass -c "irm https://astral.sh/uv/install.ps1 | iex" or winget install astral-sh.uv
    • Any platform: pip install uv
  • SSH access to this GitHub repo

1. Get your Zube API credentials

  1. Go to Zube.io → Account Settings → API Keys
  2. Create a new API key — this gives you a Client ID and a private key file (PEM)
  3. Save the private key somewhere safe (e.g. ~/.zube/private-key.pem)

2. Configure in Cursor

Add to your .cursor/mcp.json (project-level or global ~/.cursor/mcp.json):

{
  "mcpServers": {
    "zube": {
      "command": "uvx",
      "args": ["--from", "git+ssh://git@github.com/reachreporting/zube-mcp", "zube-mcp"],
      "env": {
        "ZUBE_CLIENT_ID": "your-client-id-here",
        "ZUBE_PRIVATE_KEY_PATH": "/path/to/your/private-key.pem"
      }
    }
  }
}

Replace the env values with your own Zube credentials. uvx will install the package directly from GitHub — no clone needed.

To pin a specific version, tag a release and append it to the URL:

git+ssh://git@github.com/reachreporting/zube-mcp@v0.3.0

Alternative: local install

If you prefer to clone and run locally:

git clone git@github.com:reachreporting/zube-mcp.git
cd zube-mcp
pip install -e .

Then point the Cursor config at your local path:

"args": ["--from", "/path/to/zube-mcp", "zube-mcp"]

Development

Making changes

After editing the source code, you must bump the version in pyproject.toml for Cursor to pick up changes. This is because uvx caches the built package by version — if the version hasn't changed, it serves the stale cached copy.

# 1. Edit code in zube_mcp/
# 2. Bump version in pyproject.toml (e.g. 0.2.0 → 0.3.0)
# 3. Commit and push
# 4. In Cursor: Settings → MCP → toggle zube off, then on

To verify your changes locally before restarting Cursor:

uvx --from . python3 -c "
from zube_mcp.server import mcp
import inspect
from zube_mcp.server import list_cards
print(inspect.signature(list_cards))
"

Zube API reference

Full API docs: https://zube.io/docs/api

Key concepts for adding new filters:

  • List endpoints support where[field]=value query params for filtering
  • The _build_params() helper in server.py converts a where dict into these query params automatically
  • Array filters (e.g. assignee_ids) use where[field][]=value (handled by _build_params when the value is a list)
  • Card numbers (the #12345 visible in the UI) are distinct from internal card IDs — use where[number] to filter by the visible number

Known Zube API quirks

  • get_card requires the internal card ID, not the visible card number. Use get_card_by_number to look up by the human-visible #number.
  • search_key on list endpoints is a full-text search and can be unreliable for finding cards by number.
  • List responses return items in data array with a pagination object.
  • Rate limit: 1 request/second. Short bursts are tolerated but sustained higher rates will be rejected.

Available Tools

Person & Accounts

Tool Description
get_current_person Get the authenticated user's profile
list_accounts List organizations the user belongs to
get_account Get details for a specific account

Projects

Tool Description
list_projects List projects (optionally by account)
get_project Get project details
create_project Create a new project

Workspaces (Kanban Boards)

Tool Description
list_workspaces List workspaces (optionally by project)
get_workspace Get workspace details
create_workspace Create a new workspace

Cards (Issues / PRs)

Tool Description
list_cards List cards with filters (project, workspace, sprint, epic, number, state, search)
list_project_cards List cards scoped to a project (also supports number filter)
list_triage_cards List cards in a project's triage
get_card Get full card details by internal ID
get_card_by_number Look up a card by its visible #number (e.g. 15293)
create_card Create a card
update_card Update a card
move_card Move a card to a column or triage
archive_card Archive a card

Card Comments

Tool Description
list_card_comments List comments on a card
create_card_comment Add a comment
update_card_comment Edit a comment
delete_card_comment Delete a comment

Epics

Tool Description
list_epics List epics for a project
get_epic Get epic details
create_epic Create an epic
update_epic Update an epic
list_epic_cards List cards in an epic

Sprints

Tool Description
list_sprints List sprints for a workspace
get_sprint Get sprint details
create_sprint Create a sprint
update_sprint Update a sprint

Tickets

Tool Description
list_tickets List tickets for a project
get_ticket Get ticket details
create_ticket Create a ticket
update_ticket Update a ticket

Labels & Members

Tool Description
list_labels List project labels
create_label Create a label
list_sources List connected GitHub repos
list_project_members List project members
list_account_members List account members

Architecture

zube_mcp/
  auth.py      # RS256 JWT creation for Zube's refresh token flow
  client.py    # Async HTTP client with automatic token management
  server.py    # FastMCP tool definitions (40 tools)

The auth flow:

  1. Sign a 60-second JWT with your private key
  2. Exchange it at POST /api/users/tokens for a 24-hour access token
  3. The client auto-refreshes when the token is near expiry

推荐服务器

Baidu Map

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。

官方
精选
JavaScript
Playwright MCP Server

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。

官方
精选
TypeScript
Magic Component Platform (MCP)

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。

官方
精选
本地
TypeScript
Audiense Insights MCP Server

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。

官方
精选
本地
TypeScript
VeyraX

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。

官方
精选
本地
graphlit-mcp-server

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。

官方
精选
TypeScript
Kagi MCP Server

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。

官方
精选
Python
e2b-mcp-server

e2b-mcp-server

使用 MCP 通过 e2b 运行代码。

官方
精选
Neon MCP Server

Neon MCP Server

用于与 Neon 管理 API 和数据库交互的 MCP 服务器

官方
精选
Exa MCP Server

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。

官方
精选