Great Question

io.github.GreatQuestion由 GitHub 账号 GreatQuestion 发布v1.0.0更新于 Oct 25, 2026

Run user research from any AI tool. Create studies, recruit participants, query insights.

已验证Streamable HTTP可网页运行效率与工作流知识与记忆商业与电商

概览

AI 生成的概览

Great Question 官方 MCP 服务器,让 AI 助手创建用户研究项目、招募参与者并查询研究洞察。

可以做什么
通过一句提示或对话创建、配置并启动访谈研究、问卷和原型测试。搜索客户 CRM,按属性细分并生成研究参与者名单。设置可用时间、筛选问题和激励方案。在研究库中检索过往访谈、转录稿、高亮和标签。查询已有报告并跨多个研究生成摘要。
适用场景
适合已使用 Great Question 的产品、设计和研究团队,希望不离开 AI 客户端就能执行研究流程并获取客户洞察。也适合需要快速发起研究、招募参与者或查阅过往发现的探索工作。
运行要求
远程托管端点,无需本地安装。需要一个任意套餐的 Great Question 工作区,以及支持远程 HTTP 服务器的 MCP 客户端。首次连接需通过 OAuth 2.0 登录 Great Question;未声明 API 密钥或环境变量。
安装前请注意
它并非只读:连接的智能体可创建和启动研究、招募参与者并配置激励,受工作区角色权限约束。激励方案可能涉及付费。研究数据、CRM 记录和转录稿会通过受限的 OAuth 令牌发送给 AI 客户端。

安装

在 SourceWeft 中

  1. 打开 控制台中的 Great Question,将其添加到工作区。
  2. 连接 所需的 凭据。它们会被私密存储,绝不会在公开页面上展示。
  3. 为需要使用其工具的对话启用该服务。

Web executable,通过 Streamable HTTP。 远程服务在工作区中配置后即可从网页运行时运行。

交给智能体安装

把这段提示词粘贴到 Claude Code、Codex、Cursor 等智能体,或 SourceWeft 对话中。智能体会读取该服务器的安装指南,先向你展示发布者、端点和登录方式,经你同意后再把它添加到你的客户端。密钥不会写进任何命令。

阅读 https://sourceweft.com/mcp/io.github.GreatQuestion%2Fmcp/install.md,并按其中说明把这个 MCP 服务器接入我的客户端。修改任何配置之前,先告诉我它的发布者、端点和登录方式,等我确认。不要把 API 密钥写进命令或我会提交的文件;问我密钥应从哪里获取。

查看智能体所依据的安装指南

其他 MCP 客户端

把它添加到你客户端的 mcpServers 配置中。 客户端会在首次连接时完成 OAuth 登录。

{
  "mcpServers": {
    "great-question": {
      "type": "http",
      "url": "https://greatquestion.co/api/mcp/v1"
    }
  }
}

README

Great Question MCP Server

Run user research from any AI tool. The official Model Context Protocol (MCP) server for Great Question — the AI UX research platform for product builders.

Once connected, Claude, ChatGPT, Cursor, Windsurf, Gemini CLI, and any other MCP-compatible client can create research studies, recruit participants, and query insights directly from your Great Question workspace.

"The Great Question MCP surfaces insight across studies in minutes, not days. It brings real customer patterns into discovery when decisions are happening, so we can act with confidence and build what matters most." — Matt Radel, VP of Design & UX at InvoiceCloud


What it does

This is not a read-only research lookup tool. AI agents connected to this server can actively run research workflows:

  • 🎤 Create moderated interview studies from a single prompt
  • 📝 Build multi-question surveys without leaving the conversation
  • 🧪 Set up prototype tests with screeners, scheduling, and incentives
  • 🎯 Search your customer CRM to identify the right participants for each study
  • 🔍 Query your research repository for insights, summaries, and reports across past sessions, transcripts, and tags

Who it's for

  • Product managers bringing the customer voice into PRDs and discovery work
  • UX researchers scaling research programs without scaling hands-on coordination
  • Designers referencing past usability findings while building new flows
  • Research ops teams streamlining workflows across studies and repositories
  • Founders doing product discovery without a dedicated research function

Setup

Prerequisites

  • A Great Question workspace (any plan)
  • An MCP-compatible client (Claude Desktop, Claude Code, Cursor, Windsurf, ChatGPT, Gemini CLI, or any other)

Installation

The Great Question MCP is hosted — there is nothing to install locally. You add it to your MCP client by pointing it at our endpoint:

https://greatquestion.co/api/mcp/v1

Claude Code

bash
claude mcp add --transport http great-question https://greatquestion.co/api/mcp/v1

Claude Desktop

Open Claude Desktop → Settings → Developer → Edit Config, then add:

json
{  "mcpServers": {    "great-question": {      "type": "http",      "url": "https://greatquestion.co/api/mcp/v1"    }  }}

Restart Claude Desktop. You'll be prompted to authorize via OAuth on first use.

Cursor

Cursor → Settings → MCP → Add new MCP server:

json
{  "mcpServers": {    "great-question": {      "url": "https://greatquestion.co/api/mcp/v1"    }  }}

Windsurf

Windsurf Settings → Cascade → MCP Servers → Add custom server with URL https://greatquestion.co/api/mcp/v1. Windsurf supports OAuth automatically.

ChatGPT

In ChatGPT, go to Settings → Connectors → Add connector. Use the URL https://greatquestion.co/api/mcp/v1. You'll be prompted to sign in via OAuth.

Gemini CLI

bash
gemini extensions install https://github.com/GreatQuestion/mcp

Other MCP clients

Any MCP client that supports remote (HTTP) servers can connect using the URL https://greatquestion.co/api/mcp/v1. Authentication is OAuth 2.0.

Authentication & security

  • OAuth 2.0 — credentials are never exposed to the AI client. You sign in to Great Question on first connection; the AI agent receives a scoped access token.
  • Role-based permissions — every action respects the role you have in your Great Question workspace. If you can't create studies in the UI, you can't create them via the MCP either.
  • SOC 2 Type II compliant — Great Question maintains an active SOC 2 Type II attestation; the MCP inherits the same controls as the rest of the platform.
  • Workspace isolation — your data never crosses workspace boundaries.

Example prompts

Here are some prompts that work well with the Great Question MCP. They're meant as a starting point — the server exposes the full set of research workflows, so phrasing is flexible.

  • "Create a moderated interview study about checkout abandonment, 6 participants from our enterprise customers, 30 minutes, $100 incentive."
  • "Find the 5 most-mentioned pain points across our usability tests from the last 90 days."
  • "Recruit 10 PMs from our CRM who fit our ICP for a survey on AI tool adoption."
  • "Summarize what we've heard from customers about our pricing in the last quarter."
  • "Pull every quote from our research repository where someone mentions [feature name]."

Tools exposed

The server exposes the following toolsets. Tool names and exact arguments are discovered automatically by your MCP client — this list is for orientation.

  • Studies — list, create, configure, and launch interview studies, surveys, and prototype tests
  • Participants — search your customer CRM, segment by attributes, generate participant lists for studies
  • Scheduling & screeners — configure availability, screening questions, and incentive structures
  • Repository search — full-text and semantic search across past sessions, transcripts, highlights, and tags
  • Insights & reports — query existing reports and generate summaries across multiple studies

Resources

Support

License

MIT — see LICENSE.

来源:README.md,提交 82b1bba

工具

0
该服务器需要登录,因此 SourceWeft 无法在没有账号的情况下读取其工具,也未收录。已连接账号所能看到的内容可能因账号及其权限而异。

版本历史

1
  1. v1.0.0最新Sep 16, 2026

评分

已满足适用检查项的 85%

该评分衡量此收录条目的完整程度,依据注册中心条目、端点与公开来源计算得出。它不是对服务器本身的评价,也不表示服务器安全或当前可用:可连通性是带日期的检查结果,评分也不包含安装量或用户评价。

  • 近 30 天内可连通必需25 分
  • 已声明认证方式必需10 分
  • 已列出工具15 分

    请让服务器在无凭据时响应 tools/list,或发布可在沙箱中无凭据启动的软件包。

  • 描述必需15 分
  • 图标10 分
  • 文档15 分
  • 许可证10 分