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)
Real founder decisions, lessons and signals from 100+ podcasts — searchable by AI agents.
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://echomindr.com/mcp/在终端中运行这条命令
claude mcp add --transport sse echomindr https://echomindr.com/mcp/粘贴到项目中的 .cursor/mcp.json,或 ~/.cursor/mcp.json
{
"mcpServers": {
"echomindr": {
"url": "https://echomindr.com/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": {
"echomindr": {
"type": "sse",
"url": "https://echomindr.com/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://echomindr.com/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 echomindr 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(1.0.0)于8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, AlphaGrowth
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.
Free crypto data for AI agents: TVL, fees, funding, investors, yields, token unlocks, DAO treasuries
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。
3,500+ real founder moments from 60+ podcasts — searchable by AI agents.
Each moment: a named founder, a verbatim quote, a decision taken, an outcome observed, a lesson extracted — with a timestamped link to the source. Not summaries. Not paraphrases. What actually happened.
AI agents give generic startup advice. Echomindr gives them access to what founders actually did.
Ask: *"How did founders handle their first pricing?"* Get: Kevin Hale's 10-5-20 rule, Josh Pigford charging $249/month from day one, Madhavan Ramanujam's options trick — with quotes, outcomes, and source links.
Ask: *"What did founders do when they nearly ran out of money?"* Get: Airbnb selling cereal boxes, Notion's near-collapse during COVID, Calm's years of slow growth before the breakout — directly from the founders who lived it.
# Search for founder experiences
curl "https://echomindr.com/search?q=pricing&limit=5"
curl -X POST "https://echomindr.com/situation" \ -H "Content-Type: application/json" \ -d '{"situation": "B2B SaaS founder with free pilots that won'\''t convert to paid"}'
curl "https://echomindr.com/moments/{id}"
curl "https://echomindr.com/similar/{id}?limit=5"
API docs: echomindr.com/docs
Connect via remote MCP: https://echomindr.com/mcp/
Or add to Claude Desktop (claude_desktop_config.json):
{
"mcpServers": { "echomindr": { "command": "python", "args": ["echomindr_mcp.py"], "env": { "ECHOMINDR_API_URL": "https://echomindr.com" } } } }
3 MCP tools:
search_experience — semantic search for founder stories by situationget_experience_detail — full details of a moment (quote, decision, outcome, lesson)find_similar_experiences — related founder stories by shared themeshttps://echomindr.com/llms.txt
Each moment: summary · verbatim quote · decision · outcome · lesson · stage · tags · timestamp link
To run your own instance with the sample data:
git clone https://github.com/echomindr/echomindr.git
cd echomindr pip install -r requirements.txt
python echomindr_build_db.py --sample
python echomindr_api.py
To build the full database, you need your own podcast transcriptions and Claude API key. See echomindr_extract_v2.py for the extraction pipeline.
Podcast audio → Deepgram (transcription) → Claude (extraction) → SQLite → FastAPI → MCP
The extraction pipeline turns long-form podcast interviews into structured, searchable moments. Each episode yields 8–15 moments on average. Semantic search uses BAAI/BGE-M3 embeddings (1024-dim) via sqlite-vec.
| Endpoint | Method | Description |
|---|---|---|
/search | GET | Full-text search with stage/type filters |
/situation | POST | Describe a situation, get matching experiences (vector search) |
/moments/{id} | GET | Full moment detail |
/similar/{id} | GET | Similar moments by shared tags |
/taxonomy | GET | 52 canonical situations across 10 families |
/stats | GET | Database statistics |
/llms.txt | GET | LLM-optimized API description |
/docs | GET | Swagger documentation |
MIT — the code is open source. The hosted database at echomindr.com is a managed service.
Built by Thierry — author of "The System That Learns Wins" and "Designing for Permanent Hostility".