trello-mcp
A Model Context Protocol server for Trello that connects AI agents to Trello boards with per-project configuration and local credentials.
README
trello-mcp
A Model Context Protocol server for Trello — connect AI agents to your boards with per-project configuration and credentials that stay on your machine.
Works with Cursor, Claude Desktop, VS Code Copilot, and any MCP-compatible client.
Quick setup for AI agents
Paste this one line into Cursor, Claude Code, Codex, Grok, or any agent that can fetch URLs:
Fetch https://raw.githubusercontent.com/asim-altayb/trello-mcp/main/docs/agent-setup/prompt.md and follow every step to install trello-mcp and wire it to this project.
The agent will detect your OS (Linux, macOS, or Windows), clone the server, configure credentials, write .trello-mcp.json, and register MCP — without you running terminal commands.
See AGENTS.md and docs/agent-setup/prompt.md for details.
Why use this?
Most Trello MCP integrations focus on hosted OAuth or a single global board. This server is built for developers who juggle multiple codebases and want:
- Local access — your API key + token, stored locally in
.env - Per-project boards — each repo declares which Trello boards belong to it
- Board aliases — reference
sprintorbackloginstead of raw board IDs - Agent-ready output — JSON or Markdown responses
Comparison
| Feature | trello-mcp | Atlassian Trello MCP | @delorenj/mcp-server-trello |
|---|---|---|---|
| Auth | API token (local) | Cloud OAuth | API token |
| Per-repo board profiles | .trello-mcp.json |
One workspace | Global config |
| Board aliases | Yes | No | No |
| Markdown output | Yes | N/A | Yes |
| Checklists + comments | Yes | Partial | Yes |
| Rate limiting | Built-in | Hosted | Built-in |
| Best for | Multi-project workflows | Team OAuth setups | Full-featured installs |
Quick start
1. Clone and configure credentials
git clone https://github.com/asim-altayb/trello-mcp.git
cd trello-mcp
npm install
npm run setup
npm run build
npm run setup saves TRELLO_API_KEY and TRELLO_TOKEN to .env.
Get credentials from Trello Power-Up Admin → API Key tab.
2. Pin boards to a project
From any project directory:
TRELLO_PROJECT_ROOT=/path/to/your/project npm run init-project
This writes .trello-mcp.json:
{
"name": "my-app",
"boards": [
{ "id": "abc123", "name": "Sprint Board", "alias": "sprint", "default": true },
{ "id": "def456", "name": "Backlog", "alias": "backlog" }
]
}
Safe to commit — board IDs only, no secrets.
3. Connect your AI client
Add to .mcp.json (Cursor / Claude Code) or your client's MCP config:
{
"mcpServers": {
"trello-mcp": {
"command": "node",
"args": ["/absolute/path/to/trello-mcp/dist/index.js"],
"cwd": "/absolute/path/to/your/project"
}
}
}
Restart the client after saving.
See mcp-config.example.json for a copy-paste template.
Example prompts
- "Show my project Trello overview"
- "What's on the sprint board?"
- "Search for auth bug in this project's boards"
- "Create a card: Fix webhook retry"
- "Mark the first checklist item complete on card X"
Tools
Project context
| Tool | Description |
|---|---|
trello_project_info |
Active project, configured boards, default board |
trello_project_overview |
All project boards → lists → cards |
Boards & lists
| Tool | Description |
|---|---|
trello_list_boards |
Project boards (or all with allBoards: true) |
trello_get_board |
Board with lists and cards |
trello_list_lists |
Lists on a board |
trello_get_recent_activity |
Recent board activity |
trello_add_list |
Create a list |
Cards
| Tool | Description |
|---|---|
trello_get_my_cards |
Cards assigned to you |
trello_get_card |
Full card details |
trello_get_cards_in_list |
Cards in a list |
trello_create_card |
Create a card |
trello_update_card |
Update card fields / labels |
trello_move_card |
Move card to another list |
trello_archive_card |
Archive a card |
trello_search |
Search cards/boards |
Checklists & comments
| Tool | Description |
|---|---|
trello_add_checklist |
Add checklist to card |
trello_add_checklist_item |
Add checklist item |
trello_update_checklist_item |
Update text or completion |
trello_delete_checklist_item |
Delete checklist item |
trello_add_comment |
Add comment |
trello_get_card_comments |
List comments |
trello_update_comment |
Edit comment |
trello_delete_comment |
Delete comment |
Board tools accept optional board (ID, alias, or name). When omitted, the project default is used.
Multi-project setup
One shared server install; each repo gets its own .trello-mcp.json:
| OS | Shared server + .env |
|---|---|
| Linux / macOS | ~/.local/share/trello-mcp/ |
| Windows | %LOCALAPPDATA%\trello-mcp\ |
SERVER_DIR/ → .env (credentials, once)
~/project-a/ → .trello-mcp.json (boards 1, 2)
~/project-b/ → .trello-mcp.json (boards 3, 4)
Point each project's MCP config cwd at that project. The server picks up the right boards automatically.
Security
- Never commit
.env— it contains your API token .trello-mcp.jsonis safe to commit (board IDs only)- Revoke tokens anytime: Trello → Account → Applications
Development
npm run build
npm test
npm run dev
Contributing
Issues and PRs are welcome. See CONTRIBUTING.md.
License
MIT — see LICENSE.
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。