Tiddly MCP
Connects AI agents to TiddlyWiki through the Model Context Protocol, enabling listing, searching, filtering, creating, updating, and deleting tiddlers. Defaults to read-only mode with optional write permissions.
README
Tiddly MCP
Tiddly MCP is a Model Context Protocol plugin for TiddlyWiki that lets you connect AI agents to your TiddlyWiki.
WARNING: This plugin is a work in progress.
This plugin is tested and works with:
The plugin only supports server (Node.JS-hosted) versions of TiddlyWiki.
This plugin provides the following tools to AI agents via MCP:
| Name | Description | Arguments |
|---|---|---|
list_tiddlers |
List all tiddlers or filter them using a filter expression. | filter, limit, includeSystem, includeDetails |
read_tiddler |
Read a tiddler by title. | title |
search_tiddlers |
Search for tiddlers containing specific text. | query, field, caseSensitive |
write_tiddler |
Create or update a tiddler by title. | title, text, tags, type, username |
delete_tiddler |
Deletes a tiddler by title. | title |
The plugin defaults to read only mode. The update and delete tiddler tools are off by default.
Installation
This plugin only works with server (hosted on Node.JS) TiddlyWikis.
- Drag and drop
$__plugins_rryan_tiddly-mcp.jsoninto your Node.JS-hosted TiddlyWiki and import the tiddler, or copy thetiddly-mcpfolder into your TiddlyWiki plugin path. - Open the plugin configuration and customize the port and other settings.
- Restart your server and confirm you see
[MCP] TiddlyWiki MCP Server started on port XXXXin your logs.
Authentication
For now, we recommend not exposing this MCP server to the Internet. You can use the plugin locally with Gemini CLI, Claude Code, or Claude Desktop without any security risks.
If you do expose the plugin to the Internet, we recommend:
- Set it up behind a reverse proxy, just like your Node.JS-hosted TiddlyWik is.
- Use HTTP Basic Authentication to protect access. Most clients support custom
headers that allow you to provide a
Authentication: Basic asdf1234authentication header. - Set the CORS allowed hosts to the domain name you are hosting it on.
- Use a path other than
https://example.com/mcpto avoid attackers scanning for MCP servers.
Developing
To build the plugin JSON and folder:
pnpm install
pnpm build:folder
To run the tests:
pnpm run test
Acknowledgements
- This plugin was initially authored by Claude Code and Gemini CLI.
- Thanks to https://github.com/tiddly-gittly/Modern.TiddlyDev for the nice TypeScript starter kit for TiddlyWiki plugin development.
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。