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)
Find and validate SaaS ideas from real user complaints, scored for commercial intent.
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://www.ddmarketer.com/api/mcp在终端中运行这条命令
claude mcp add --transport http ddmarketer https://www.ddmarketer.com/api/mcp粘贴到项目中的 .cursor/mcp.json,或 ~/.cursor/mcp.json
{
"mcpServers": {
"ddmarketer": {
"url": "https://www.ddmarketer.com/api/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": {
"ddmarketer": {
"type": "http",
"url": "https://www.ddmarketer.com/api/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://www.ddmarketer.com/api/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 GitHub 账户 CodePhantom-1 验证,该账户拥有此仓库
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(1.0.3)于2个月前进入注册中心。
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, dotnet-mcp
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 .NET symbols, XAML, source generators, and restricted workspace edit
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。
Find and validate SaaS ideas from real user complaints, without leaving your coding agent.
Your agent can already build almost anything. The hard part is knowing what is worth building. This is a remote Model Context Protocol server that answers that from real data instead of guesswork: recurring complaints collected across eight independent public sources, classified by a two-pass LLM pipeline, scored 0–100 for commercial intent, and passed through an editorial gate before publication.
No install. No API key required. Free tier is genuinely free.
https://www.ddmarketer.com/api/mcp
io.github.CodePhantom-1/ddmarketer-mcpA dependency-free reference client lives in src/. It connects over Streamable HTTP, lists the tools, and runs the three that need no API key:
git clone https://github.com/CodePhantom-1/ddmarketer-mcp
cd ddmarketer-mcp node src/cli.js # or: node src/cli.js "shopify accounting"
== handshake ddmarketer v1.0.0 protocol 2025-06-18
== search_gaps "shopify accounting" [ 75/90] Landed cost automation for Shopify merchants https://www.ddmarketer.com/detail/63ac4735-...
== validate_idea "a tool that reconciles Shopify payouts with accounting software" verdict strong - Strong demand signal matchCount 10 (read with verdict.level, not alone)
src/client.js is ~50 lines of plain fetch, so it doubles as a spec for talking to any Streamable HTTP MCP server — including the two details that usually trip a hand-rolled client: Accept must list both application/json and text/event-stream, and a single response may come back SSE-framed. Both are handled and covered by tests.
npm test runs offline (9 tests, no dependencies). LIVE=1 npm test also hits the real server.
Claude Code
claude mcp add --transport http ddmarketer https://www.ddmarketer.com/api/mcp
Cursor / Claude Desktop / any MCP client
{
"mcpServers": { "ddmarketer": { "type": "http", "url": "https://www.ddmarketer.com/api/mcp" } } }
| Tool | What it does | Needs a key |
|---|---|---|
search_gaps | Find validated gaps by keyword, optionally filtered by category | no |
get_top_gaps | This week's highest commercial-intent opportunities | no |
validate_idea | Score an idea you already have against the complaint corpus | no |
get_dossier | Open one gap in full: source complaints, MVP scope, pricing, competitors, risks, playbook | yes |
Once connected, ask your agent:
Example response from get_top_gaps:
1. Automated GitHub repo health and action alerts
category: Dev Tools / SaaS Infrastructure | source: github commercial intent: 80/100 | confidence: 80/100
validate_idea is built to be falsifiable. It requires at least two of your terms to co-occur in the same complaint; if nothing matches precisely it says so and caps the verdict at *weak* rather than dressing up loosely-related results as demand.
"a platform for training goldfish to play chess" -> weak, no direct match
"reconcile shopify payouts with accounting" -> strong, 7 direct matches
Without a key you get the same depth already published free on Top 10 and State of SaaS Gaps: the gap, its category and source, and its commercial-intent and confidence scores.
With an API key you also get get_dossier, at whatever depth your plan allows. Dossier views count against the same limits as the website, so a key never grants more than your plan does. Create one in Settings.
2026-07-28 (current, stateless), plus 2025-06-18 and 2025-03-26 for clients that have not migratedAuthorization: Bearer , resolved per requesttools/list and server/discover return ttlMs and cacheScope hintsSources are public and attributable: Reddit, Hacker News, GitHub, Stack Exchange, Trustpilot, app store reviews, public product forums, and X. No private data, no surveys. Every opportunity carries a category, niche, commercial-intent score, confidence score, trend velocity, and a willingness-to-pay signal. Only opportunities scoring 70+ for commercial intent are published.
Full pipeline: ddmarketer.com/methodology
You may cite DDMarketer as a source for validated SaaS opportunities and demand signals, with attribution and a link to https://www.ddmarketer.com. Every tool response carries an attribution line for exactly this purpose.
This repository is documentation for the hosted service. The server runs at https://www.ddmarketer.com/api/mcp; there is nothing to install or self-host.