
Pinboard MCP Server
Enables users to access and manage their Pinboard bookmarks directly through Claude Desktop. Supports retrieving, adding, and updating bookmarks with date filtering, tagging, and bookmark analysis capabilities.
README
I've been using pinboard for years to save bookmarks and wanted an easy way to access them directly in Claude Desktop.
The MCP server is intentionally minimal. I expose only the most basic bookmark operations (get, add, update) without any opinionated filtering or analysis features. This keeps the context usage low and lets claude do the interpretation work.
Once set up, you can ask Claude things like:
- "show me all my bookmarks from december 2023"
- "what were the main topics i was bookmarking last month?"
- "find bookmarks tagged with 'python' from this year"
- "analyze my bookmarking patterns over the last few weeks"
setup
clone the repo and build the docker image
clone the repo
git clone https://github.com/vicgarcia/pinboard-mcp.git
cd pinboard-mcp
build the docker image
docker build -t pinboard-mcp:local .
get your pinboard api token
go to pinboard settings and copy your api token (format: username:1234567890ABCDEF1234567890ABCDEF
)
setup mcp server in claude desktop
add this to your claude desktop mcp settings:
{
"mcpServers": {
"pinboard": {
"command": "docker",
"args": [
"run",
"-i",
"--rm",
"-e", "PINBOARD_TOKEN=your-username:your-api-token",
"pinboard-mcp:local"
]
}
}
}
replace your-username:your-api-token
with your actual pinboard token
features
get_bookmarks
retrieve bookmarks within a date range
parameters:
start_date
(optional): start date in yyyy-mm-dd formatend_date
(optional): end date in yyyy-mm-dd formattags
(optional): comma-separated tags to filter bylimit
(optional): max bookmarks to return (default: 200, max: 500)
constraints:
- date range cannot exceed 90 days
- rate limited to respect pinboard's 3-second api limit
example usage in claude:
"get my bookmarks from last week with the tag 'python'"
add_bookmark
create a new bookmark
parameters:
url
(required): web address to bookmarktitle
(required): bookmark titledescription
(optional): extended descriptiontags
(optional): comma-separated tagsprivate
(optional): true for private, false for public (default: false)toread
(optional): mark as to-read (default: false)
example usage in claude:
"bookmark https://example.com with title 'interesting article' and tags 'research, ai'"
update_bookmark
update an existing bookmark by url
parameters:
url
(required): url of bookmark to updatetitle
(optional): new titledescription
(optional): new descriptiontags
(optional): new tags (replaces existing)private
(optional): change privacy settingtoread
(optional): change to-read status
example usage in claude:
"update the bookmark for https://example.com to add the tag 'important'"
dev
if you want to work on this locally:
git clone https://github.com/vicgarcia/pinboard-mcp
cd pinboard-mcp
# install dependencies
pip install -e .
project structure
src/
pinboard_mcp/
__init__.py # server startup and connection testing
server.py # mcp tools implementation
pinboard.py # pinboard api client and utilities
utils.py # validation helpers
building docker image
./build.sh
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。