KnowUsWell MCP Server
Hosts the KnowUsWell friendship quiz in conversations, scoring 16 behavior questions into four-letter roles and combining two roles into one of 16 pair types with shared result links.
README
knowuswell-mcp
MCP (Model Context Protocol) server for KnowUsWell — the friendship quiz that needs two people.
Let your AI assistant host the quiz right in the conversation: it asks you 16 quick behaviour questions, scores your four-letter role, and — once your friend has answered their half — combines both codes into one of 16 pair types, a card that belongs to the friendship rather than to either person.
Tools
| Tool | What it does |
|---|---|
get_quiz_questions |
The 16 behaviour questions (true = "more like me") |
score_quiz |
Answers → four-letter role code + full role card + link |
combine_pair |
Two role codes → one of 16 pair types + shared result link |
get_conversation_questions |
50 read-aloud "how well do you know me" questions, easy to brutal |
Scoring is deterministic and identical to knowuswell.com — the same answers always produce the same code, and ties break the same way. Result links open the shared card page on the site, where you can collect all 16 types.
Install
Claude Desktop / Claude Code
{
"mcpServers": {
"knowuswell": {
"command": "npx",
"args": ["-y", "knowuswell-mcp"]
}
}
}
(Published on npm as knowuswell-mcp.)
Or clone and run locally:
git clone https://github.com/ppop123/knowuswell-mcp.git
cd knowuswell-mcp && npm install
{
"mcpServers": {
"knowuswell": {
"command": "node",
"args": ["/path/to/knowuswell-mcp/index.js"]
}
}
}
Try it
"Host the KnowUsWell friendship quiz for me."
The assistant asks the 16 questions, tells you your role (say, ESPR — The Cat: appears on your terms, still yours), and gives you a link to send your friend. When their code comes back:
"We got LOCT and ESPR — what are we?"
…it combines them into your pair type (The Anchor and the Kite, in this case) with what works, what to watch, and a shared card page for the two of you.
Privacy
The server is a pure local calculator: questions, scoring tables and card text ship in data.json. Nothing you answer is sent anywhere — the only network activity is you choosing to open a result link. (The site works the same way: answers travel inside the URL fragment, which browsers never transmit.)
About
Built by KnowUsWell. The quiz is free to play on the site — no account, nothing stored. One optional extra exists: a $1.99 illustrated art card of your pair type, drawn from your own photos and clearly labelled AI-generated.
MIT licensed.
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。