Spec Driven Development
MCP Server that facilitates spec-driven development workflows, not just Vibe Coding.
作者提供的说明(英文原文)
Push CSV, JSON, JSONL or Parquet from your agent and get a link a human can explore.
这三项永远不会合并成一个分数:意见、测量和人气是不同的东西。
我们还没有为它写通俗的介绍。下面是我们从注册中心读取到的事实。
选择你使用的 AI 应用,然后复制设置。
你需要
你用的是哪个应用?
在上方选择你的应用。我们会记住它,下次按钮会显示“安装到 …”。
适用于网页版、桌面版和手机版 Claude:
https://rows.page/mcp在终端中运行这条命令
claude mcp add --transport http rows https://rows.page/mcp粘贴到项目中的 .cursor/mcp.json,或 ~/.cursor/mcp.json
{
"mcpServers": {
"rows": {
"url": "https://rows.page/mcp"
}
}
}浏览器会先询问是否打开应用,然后应用会显示设置并请你确认。密钥的值会留空,由你自己填写。
粘贴到项目中的 .vscode/mcp.json
{
"servers": {
"rows": {
"type": "http",
"url": "https://rows.page/mcp"
}
}
}浏览器会先询问是否打开应用,然后应用会显示设置并请你确认。密钥的值会留空,由你自己填写。
ChatGPT 只能连接互联网上的服务器:
https://rows.page/mcp菜单名称可能因套餐和版本不同而略有差异。
你的选择通过 Cookie 保存在这台设备上,无需账户。
这些设置是根据作者在注册中心的条目生成的。如果 README 的写法不同,请以 README 为准。
我们还没有读取这个服务器的工具列表;我们会轮流检查各个端点。
我们检查的是任何人都能看到的信息;我们没有阅读代码。不适用的检查项不计入总数。 · 检查于 2026年11月6日
注册名称已通过域名 rows.page 验证(注册中心会核实域名所有权)
我们还没有从 GitHub 读取这个仓库
我们还没有从 GitHub 读取这个仓库
尚未检查;我们会轮流检查各个端点
尚未检查;我们会轮流检查各个端点
尚未检查;我们会在检查端点时读取工具列表
与注册中心里这个分类中所有活跃的服务器比较。只有当发布者拥有该仓库时,星标数才计入。
它在 1 项适用的基础检查中通过了 1 项。我们没有阅读代码。在添加密钥或令牌之前,请先看看这个服务器要求哪些权限,并使用只具备所需权限的密钥。
注册中心没有列出必需的密钥,不过有些服务器会在首次使用时要求你登录。
它在作者的云端运行。通过互联网连接,无需安装,但你发送的内容会经过作者的服务器。
GitHub 仓库最后更新于–,最新版本(1.0.0)于28天前进入注册中心。
支持 MCP 的应用,例如 Claude Desktop、Claude Code、Cursor 和 VS Code。它通过 URL 连接,所以接受远程 MCP 服务器的应用也可以使用,例如开启开发者模式的 ChatGPT。
同一分类中有:Spec Driven Development, Remote Desktop Commander, Safe Fetch
同一分类中其他作者的服务器,按星标数排序。
如何看卡片:★ GitHub 星标i三大支柱:评价 · 检查 · 使用情况i更新 = 最新版本发布到注册中心的时间i
MCP Server that facilitates spec-driven development workflows, not just Vibe Coding.
Remote Desktop Commander: use files, terminal and processes on your own computers from any AI app.
MCP server for fetching URLs, safe against SSRF, DNS rebinding, and redirect-to-internal attacks.
按原文从作者的仓库复制。PopNx MCP 没有审查或背书其内容。HTML、图片和脚本已被移除。链接会离开本站。
rows.page is a place for your rows. Your agent pushes CSV, TSV, JSON, JSONL or Parquet with one tool call and gets back a link a human can explore: filter, chart columns, run SQL, read pinned findings, and see what changed since the last push. Viewing is free, no signup needed. Anonymous datasets live 7 days unless someone keeps them.
The server is remote only: https://rows.page/mcp, streamable HTTP, stateless JSON-RPC. There is nothing to install. A GET on the endpoint returns the server card.
Claude Code
claude mcp add --transport http rows https://rows.page/mcp
Claude Desktop (claude_desktop_config.json, needs Node.js)
{
"mcpServers": { "rows": { "command": "npx", "args": ["-y", "mcp-remote", "https://rows.page/mcp"] } } }
Cursor (.cursor/mcp.json)
{
"mcpServers": { "rows": { "url": "https://rows.page/mcp" } } }
VS Code (.vscode/mcp.json)
{
"servers": { "rows": { "type": "http", "url": "https://rows.page/mcp" } } }
Windsurf (mcp_config.json)
{
"mcpServers": { "rows": { "serverUrl": "https://rows.page/mcp" } } }
Codex CLI (~/.codex/config.toml)
[mcp_servers.rows]
url = "https://rows.page/mcp"
Gemini CLI (~/.gemini/settings.json)
{
"mcpServers": { "rows": { "url": "https://rows.page/mcp", "type": "http" } } }
Anonymous pushes work. An API key (created at https://rows.page/account) pushes into your account, unlocks list_datasets, and lets you re-push by a stable key. Send it as Authorization: Bearer rpk_... or append ?apiKey=rpk_... to the URL.
claude mcp add --transport http rows https://rows.page/mcp --header "Authorization: Bearer rpk_..."
| Tool | What it does |
|---|---|
push | Push csv, tsv, json, jsonl, rows (array of objects) or a url. Optional title, summary, key, id_column, findings[] and provenance (source, run, commit). Pass dataset + token for a new version. Returns the link plus next steps for your human. |
pin_finding | Pin a saved view (title, note, q filter, sort, cols or SQL) to a dataset. The page opens on its findings. |
get_dataset | Meta, versions, findings, columns and the first 20 rows. |
list_datasets | Your datasets, newest first. Needs an API key. |
delete_dataset | Delete a dataset you own. Its URL stops working immediately. |
get_limits | Your plan, its caps and usage. |
Every push is an immutable version. Identical bytes return unchanged: true. Re-pushes report rows new, gone and changed (set id_column so changed rows can be matched) plus a diff URL. Finding filters use the rows.page query language (status:failed latency_ms:>2000, (country:NL OR country:BE) amount:>500) or SQL starting with SELECT.
Push results.csv to rows.page titled "Eval run 412", pin the failed rows as a finding with a one-line note on what they have in common, and give me the link.
Query the last 24h of orders, push them to rows.page with key nightly-orders and id_column order_id, then tell me how many rows changed since yesterday.
Fetch https://example.com/export.parquet, push it to rows.page, pin a finding for amount:>500 sorted by -amount, and hand me the owner link.
Anonymous: 100 MB per file (16 MB inline per MCP push, use url for bigger files), deleted 7 days after creation unless kept, 50 pushes and 1 GB per day per IP, 60 MCP calls per minute. Free, Pro and Max plans: https://rows.page/pricing.
page.rows/rowsMade by Bitgate. MIT licensed.