productboard-mcp
Enables AI assistants to interact with your Productboard workspace, allowing features, products, notes, objectives, key results, and releases management.
README
productboard-mcp
An MCP (Model Context Protocol) server for the Productboard API. Enables AI assistants (Claude, Cursor, etc.) to interact with your Productboard workspace.
Tools
| Tool | Description |
|---|---|
pb_feature_list |
List features with filtering |
pb_feature_get |
Get a feature by ID |
pb_feature_create |
Create a new feature |
pb_feature_update |
Update an existing feature |
pb_feature_delete |
Delete a feature |
pb_product_list |
List products |
pb_product_create |
Create a product |
pb_product_hierarchy |
Get full product hierarchy |
pb_note_list |
List customer notes |
pb_note_create |
Create a customer note |
pb_objective_list |
List objectives |
pb_objective_create |
Create an objective |
pb_objective_update |
Update an objective |
pb_keyresult_list |
List key results |
pb_keyresult_create |
Create a key result |
pb_keyresult_update |
Update a key result |
pb_release_list |
List releases |
pb_release_create |
Create a release |
pb_release_update |
Update a release |
pb_release_status_update |
Update release status |
pb_release_timeline |
Get release timeline |
Installation
Option 1: One-click install (.mcpb bundle) — Recommended
Download the latest .mcpb file from the Releases page and drag it into Claude Desktop (Developer → Extensions → Install), or double-click it in a compatible MCP client.
The bundle is self-contained — no cloning or building required.
After installing, set your PRODUCTBOARD_API_TOKEN in the extension settings.
Option 2: Local install (manual)
# 1. Clone the repo
git clone https://github.com/Enreign/productboard-mcp.git
cd productboard-mcp
# 2. Install dependencies and build
npm install --include=dev
npm run build
Then add to your claude_desktop_config.json:
{
"mcpServers": {
"productboard": {
"command": "node",
"args": ["/absolute/path/to/productboard-mcp/dist/index.js"],
"env": {
"PRODUCTBOARD_API_TOKEN": "your-api-token-here",
"LOG_LEVEL": "error"
}
}
}
}
Important: Set
LOG_LEVELtoerror(notinfo). MCP uses stdio for communication — info-level logs printed to stdout will interfere with the protocol and cause the server to lock up.
Option 3: npx
⚠️ Coming soon — not yet published to npm. Use Option 1 or 2 above.
{
"mcpServers": {
"productboard": {
"command": "npx",
"args": ["-y", "@enreign/productboard-mcp"],
"env": {
"PRODUCTBOARD_API_TOKEN": "your-api-token-here",
"LOG_LEVEL": "error"
}
}
}
}
Getting a Productboard API Token
- Log in to your Productboard workspace
- Go to Profile & Settings → API Access
- Click Generate API key and copy the token
Configuration
Required
| Variable | Description |
|---|---|
PRODUCTBOARD_API_TOKEN |
Your Productboard API token (Bearer auth) |
Optional
| Variable | Default | Description |
|---|---|---|
PRODUCTBOARD_AUTH_TYPE |
bearer |
Auth type: bearer or oauth2 |
PRODUCTBOARD_API_BASE_URL |
https://api.productboard.com/v2 |
API base URL |
PRODUCTBOARD_API_TIMEOUT |
10000 |
API request timeout (ms) |
API_RETRY_ATTEMPTS |
3 |
Number of retry attempts |
API_RETRY_DELAY |
1000 |
Delay between retries (ms) |
RATE_LIMIT_GLOBAL |
100 |
Max requests per window |
RATE_LIMIT_WINDOW_MS |
60000 |
Rate limit window (ms) |
CACHE_ENABLED |
false |
Enable response caching |
CACHE_TTL |
300 |
Cache TTL (seconds) |
LOG_LEVEL |
info |
Log level: trace, debug, info, warn, error, fatal |
Note for MCP clients: Always set
LOG_LEVEL=errorwhen using with Claude Desktop, Cursor, or any stdio-based MCP client. Higher log levels write to stdout and will break the MCP protocol.
OAuth2 (optional)
| Variable | Description |
|---|---|
PRODUCTBOARD_OAUTH_CLIENT_ID |
OAuth2 client ID |
PRODUCTBOARD_OAUTH_CLIENT_SECRET |
OAuth2 client secret |
PRODUCTBOARD_OAUTH_REDIRECT_URI |
OAuth2 redirect URI |
Troubleshooting
"MCP server locks up / produces error logs"
→ Add "LOG_LEVEL": "error" to the env block in your MCP config. Info logs written to stdout interfere with the stdio transport.
"npx fails / package not found"
→ The package is not yet published to npm. Use the .mcpb bundle or local install above.
"command not found after local build"
→ Point args at the full absolute path to dist/index.js, not the productboard-mcp.js wrapper.
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。