Plate MCP
Lightweight project management for teams and AI agents.
README
Plate MCP Server
Plate is a lightweight project and task manager built for automation. Plan work, track tasks, and keep projects moving forward with people and AI agents working side by side. This MCP server lets you manage workspaces, projects, and tasks directly from Claude, Cursor, Windsurf, and any MCP-compatible AI client.
Endpoint: https://plate.to/mcp
Setup
Add this to your MCP configuration:
{
"mcpServers": {
"plate": {
"type": "http",
"url": "https://plate.to/mcp"
}
}
}
| Client | Config location |
|---|---|
| Claude Code | .mcp.json in project root, or ~/.claude/.mcp.json globally |
| Claude Desktop | ~/Library/Application Support/Claude/claude_desktop_config.json |
| Cursor | Settings → Cursor Settings → MCP |
| Windsurf | Windsurf Settings → MCP Servers |
After adding the server, your client will open a browser window to authorize via your Plate account. No API key setup required — OAuth 2.0 with PKCE handles everything automatically.
Tools
All tools return structured output (JSON Schema defined via outputSchema), compatible with OpenAI, ChatGPT Desktop, and any client that supports the MCP structured content spec.
Read tools (always available)
| Tool | Description |
|---|---|
list_workspaces |
List all workspaces you belong to |
list_projects |
List projects in a workspace |
list_sections |
List sections in a project |
list_tasks |
List tasks in a project (up to 100 by default, configurable with limit), or find by public number (workspaceId + number) |
get_task |
Get full task details — accepts internal ID, prefixed ref (SCD-426), or workspaceId + number |
list_statuses |
List workflow statuses for a workspace |
list_members |
List workspace members |
Write tools (require write scope)
| Tool | Description |
|---|---|
create_task |
Create a task in a project section (use update_task for existing tasks) |
update_task |
Update task name, status, assignee, section, or description |
complete_task |
Mark a task as done or reopen it |
delete_task |
Permanently delete a task |
create_project |
Create a new project — returns existing project if name already exists (created: false) |
update_project |
Rename a project or update its description |
create_section |
Add a section to a project — returns existing section if name already exists (created: false) |
update_section |
Rename a section |
create_comment |
Post a comment on a task |
delete_comment |
Delete a comment |
Batch write tools (require write scope)
Use batch tools when you need to create, update, or delete multiple items at once. They execute atomically and reduce the number of confirmation prompts in AI clients that ask once per tool call.
| Tool | Limit | Description |
|---|---|---|
create_tasks |
50 | Create multiple tasks in a project at once |
update_tasks |
50 | Update multiple tasks at once |
complete_tasks |
100 | Complete (or reopen) multiple tasks at once |
delete_tasks |
50 | Permanently delete multiple tasks at once |
create_sections |
30 | Create multiple sections in a project at once |
update_sections |
30 | Rename multiple sections at once |
create_comments |
50 | Add multiple comments to tasks at once |
delete_comments |
50 | Delete multiple comments at once |
Authentication
Uses OAuth 2.0 with PKCE. On first use, your client opens a browser to sign into your Plate account. You choose a scope (read or read write) and approve access. Tokens refresh automatically — you won't need to re-authorize unless the refresh token expires (30 days).
To revoke access: Workspace Settings → Apps in Plate.
Documentation
Full tool reference and parameter details: plate.to/docs/mcp
Questions: hello@plate.to
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。