MCP Server Trello
Enables AI agents to manage Trello boards, lists, cards, checklists, and workspace navigation through 23 typed MCP tools, with rate limiting and validation.
README
MCP Server Trello
Professional Model Context Protocol (MCP) server and local CLI for Trello. It exposes 23 typed tools for boards, lists, cards, checklists, activity, and workspace navigation, with Trello-aware rate limiting and validation built in.
This repository is maintained as ftaricano/mcp-server-trello. It is derived from MIT-licensed original work by Jarad DeLorenzo, with original copyright preserved and current ownership/provenance documented in NOTICE.md.
Highlights
- 23 MCP tools for cards, lists, boards, workspaces, checklists, attachments, and activity.
- Multi-board and workspace support with persisted active board/workspace state.
- Trello API rate limiting for the documented per-key and per-token limits.
- Zod input validation and TypeScript declarations in the published package.
trelloCLI for local agent workflows.- CI on Node.js 20 and 22 for lint, typecheck, tests, build, and
npm pack --dry-run.
Requirements
- Node.js 20 or newer.
- A Trello API key from https://trello.com/app-key.
- A Trello token created from that API key.
Install
Use the package directly from npm-compatible MCP clients:
{
"mcpServers": {
"trello": {
"command": "npx",
"args": ["-y", "@ftaricano/mcp-server-trello"],
"env": {
"TRELLO_API_KEY": "replace_with_trello_api_key",
"TRELLO_TOKEN": "replace_with_trello_token"
}
}
}
}
With pnpx:
{
"mcpServers": {
"trello": {
"command": "pnpx",
"args": ["@ftaricano/mcp-server-trello"],
"env": {
"TRELLO_API_KEY": "replace_with_trello_api_key",
"TRELLO_TOKEN": "replace_with_trello_token"
}
}
}
}
Global install is also supported:
npm install -g @ftaricano/mcp-server-trello
mcp-server-trello
Trello Token
Create a Trello token by visiting this URL after replacing the app name and API key:
https://trello.com/1/authorize?expiration=never&name=YOUR_APP_NAME&scope=read,write&response_type=token&key=YOUR_API_KEY
Use expiration=30days or another shorter value when your operating model can tolerate periodic token renewal. Treat the token like a password.
Configuration
The server reads credentials from environment variables:
TRELLO_API_KEY=replace_with_trello_api_key
TRELLO_TOKEN=replace_with_trello_token
# Optional defaults. These can also be changed at runtime by MCP tools.
TRELLO_BOARD_ID=
TRELLO_WORKSPACE_ID=
The CLI also loads a .env file from the current working directory and can read credentials from the macOS Keychain when TRELLO_KEYCHAIN_PREFIX is set:
security add-generic-password -s TRELLO_API_KEY -a "$USER" -w "replace_with_trello_api_key"
security add-generic-password -s TRELLO_TOKEN -a "$USER" -w "replace_with_trello_token"
export TRELLO_KEYCHAIN_PREFIX=TRELLO
trello list-boards --md
Credential resolution order for the CLI:
- Process environment.
.envfile in the current working directory.- macOS Keychain entries named
<PREFIX>_API_KEY,<PREFIX>_TOKEN, and optionally<PREFIX>_BOARD_ID.
Tools
Card tools:
add_card_to_listupdate_card_detailsarchive_cardmove_cardget_cardget_cards_by_list_idget_my_cardsattach_image_to_card
List, board, and workspace tools:
get_listsadd_list_to_boardarchive_listlist_boardsset_active_boardlist_workspacesset_active_workspacelist_boards_in_workspaceget_active_board_infoget_recent_activity
Checklist tools:
get_checklist_itemsadd_checklist_itemfind_checklist_items_by_descriptionget_acceptance_criteriaget_checklist_by_name
Most board-scoped tools accept an optional boardId. If omitted, the configured or persisted active board is used.
CLI
The package installs a trello binary for local agent and terminal workflows.
trello list-boards --md
trello set-board <boardId>
trello active-board --md
trello lists --md
trello board labels --md
trello board members --md
trello card add <listId> "Task name" --desc "details" --due 2026-05-01T12:00:00Z
trello card update <cardId> --name "Renamed" --done
trello card move <cardId> <listId>
trello card get <cardId> --md
trello card archive <cardId>
trello card comment <cardId> "Status update"
trello card attach <cardId> https://example.com/cover.png --name "Cover"
trello card assign <cardId> <memberId>
trello card unassign <cardId> <memberId>
trello cards mine --md
trello cards list <listId> --md
Default output is JSON for agent consumption. Add --md for markdown output. Exit codes are 0 for success, 1 for missing config or validation errors, and 2 for Trello API errors.
Date Formats
dueDate: full ISO 8601 timestamp, for example2026-05-01T12:00:00Z.start: date only, for example2026-05-01.
Rate Limiting
Requests are queued through a token bucket limiter for Trello's published API limits:
- 300 requests per 10 seconds per API key.
- 100 requests per 10 seconds per token.
Development
git clone https://github.com/ftaricano/mcp-server-trello.git
cd mcp-server-trello
SKIP_PREPARE=true npm install
cp .env.example .env
Useful commands:
npm run lint
npm run typecheck
npm test
npm run build
npm pack --dry-run
Security
Do not commit real Trello credentials, .env files, local logs, generated databases, or machine-specific agent configuration. See SECURITY.md for supported reporting channels and credential rotation guidance.
Contributing
Before opening a PR, run:
npm run lint
npm run typecheck
npm test
npm run build
npm pack --dry-run
Keep changes focused and include tests for behavior changes.
Provenance
This project started as a fork of delorenj/mcp-server-trello. Credit for the original MCP server scaffolding, base Trello API integration, and initial tool surface belongs to Jarad DeLorenzo.
This repository is maintained under ftaricano/mcp-server-trello. Additions include the trello CLI binary, macOS Keychain credential resolution, expanded CLI card/board/list coverage, 152 unit tests, CI on Node.js 20/22, package metadata ownership, security documentation, and npm pack hardening.
See NOTICE.md for the provenance statement.
Acknowledgments
- Original MCP server by Jarad DeLorenzo (delorenj/mcp-server-trello)
- Built with the Model Context Protocol SDK
- Uses the Trello REST API
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。