BigQuery MCP by InsightfulPipe

com.insightfulpipe由 insightfulpipe.com 发布v1.0.0更新于 Nov 4, 2026

Hosted BigQuery MCP server for Claude, ChatGPT and Cursor, with each action switched per account.

已验证Streamable HTTP可网页运行数据库数据与分析

概览

AI 生成的概览

托管的 BigQuery MCP 服务器，让 AI 助手列出数据集与表、查看表结构，并在已连接的 Google Cloud 项目上执行 SQL。

可以做什么
  • 列出已连接 Google Cloud 项目中的数据集，以及某个数据集内的表和视图
  • 查看表或视图的字段结构，以及行数、大小、分区、过期时间等完整元数据
  • 执行 Standard SQL 查询并返回结果行
  • 按已连接账号设置只读或读写，并逐个开启或关闭具体操作
适用场景
适合希望让助手直接探索 BigQuery 数据、用 SQL 回答问题而不必自建连接器的场景，尤其是需要按账号控制代理可调用哪些操作时。
运行要求
需要在 Claude、ChatGPT（开发者模式）、Claude Code 或 Cursor 中添加远程 MCP 端点。需通过 OAuth 登录 InsightfulPipe，并准备一个 Google Cloud 项目及服务账号，其角色决定只读或读写权限。7 天试用后需付费订阅。
安装前请注意
run_query 会执行 SQL，在读写连接且服务账号具备写权限时可能修改数据；破坏性操作默认关闭，但管理员可以开启。查询结果和表结构数据会经过 InsightfulPipe 托管服务。使用需订阅，起价每月 29.99 美元。

安装

在 SourceWeft 中

  1. 打开 控制台中的 BigQuery MCP by InsightfulPipe，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/com.insightfulpipe%2Fbigquery/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": {
    "bigquery-by-insightfulpipe": {
      "type": "http",
      "url": "https://bigquery.insightfulmcp.com/"
    }
  }
}

README

BigQuery MCP Server by Insightful Pipe

[MCP Compatible] [Insightful Pipe] [License]

Connect BigQuery to AI assistants: datasets, schemas and SQL on Google Cloud.

Part of the Insightful Pipe MCP Server Collection — use BigQuery from Claude, ChatGPT, Cursor, and other AI assistants through the Model Context Protocol (MCP).

[BigQuery MCP Server]

MCP Server URL

https://bigquery.insightfulmcp.com/

What is BigQuery MCP?

BigQuery MCP is a remote Model Context Protocol server hosted by InsightfulPipe. Query BigQuery datasets, list tables, inspect schemas, and run SQL on your connected Google Cloud projects. The roles you grant the service account in GCP decide whether the connection is read-only or read & write.

Installation

Claude

  1. Copy the MCP Server URL: https://bigquery.insightfulmcp.com/
  2. Open Claude Connectors Settings
  3. Scroll to the bottom and click Add custom connector
  4. Paste the URL and click Add
  5. Click Connect and sign in to InsightfulPipe if asked. Claude then lists the connector as connected.

ChatGPT

Custom MCP servers are added through ChatGPT's Developer mode. Availability depends on your ChatGPT plan, and workspace admins may need to allow it.

  1. Turn on Developer mode in ChatGPT settings
  2. Create a new app for a remote MCP server and paste the URL: https://bigquery.insightfulmcp.com/
  3. Sign in to InsightfulPipe if asked. The connection finishes as soon as you are signed in.

See OpenAI's guide: Developer mode and MCP apps in ChatGPT

Claude Code

bash
claude mcp add --transport http bigquery https://bigquery.insightfulmcp.com/

Cursor

Add the server to ~/.cursor/mcp.json (all projects) or .cursor/mcp.json (one project):

json
{  "mcpServers": {    "bigquery": {      "url": "https://bigquery.insightfulmcp.com/"    }  }}

When Cursor shows Needs authentication, click Connect and sign in to InsightfulPipe if asked.

Available Actions

5 actions: 4 read, 1 write.

Read Actions (4)

ActionDescription
get_table_metadataGet full metadata for a table or view: schema, row count, size, partitioning, expiration
get_table_schemaGet the schema (column names + types) for a table or view
list_datasetsList all datasets in the connected Google Cloud project
list_tablesList all tables and views in a dataset

Write Actions (1)

ActionDescription
run_queryExecute a Standard SQL query and return rows

Control What Your AI Can Do

You decide what AI agents can do with each connected account:

  • Turn individual actions on or off for every connected account, so agents only see the actions you allow.
  • Connect as Read-only or Read & Write. A read-only connection can only enable read actions.
  • Destructive actions stay off by default. Actions such as deletes are disabled until an admin enables them.
  • Team access per account. Restricted team members only use the accounts they are granted, with the read actions enabled on them.

Usage Examples

"List the datasets in my project"
"Show the schema of the events table"
"Run a query for daily sessions over the last 30 days"

Pricing

The BigQuery MCP server is included in every InsightfulPipe plan, together with all other MCP servers and the CLI. Plans start at $29.99/month, and you can try it for 7 days. See insightfulpipe.com/pricing for current plans.

Explore More MCP Servers by Insightful Pipe

Visit insightfulpipe.com/mcp-servers to discover our full collection of MCP servers.

View All MCP Servers →

Resources

Support

Insightful Pipe — AI-powered marketing analytics through MCP servers. Explore all integrations →

License

MIT License - see LICENSE for details.

来源：README.md，提交 9916324

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Oct 21, 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