正在打开每日关注
正在打开每日关注
通过 Markdown、llms.txt、站点地图和结构化 API,让 AI 助手更稳定地发现与读取创业公司数据。
It works! AI labs started indexing, training, and serving my new *.md files ✨ Right after pushing to prod, I asked ChatGPT to check {{startup_markdown_URL}}, and it couldn't because of "cache missed" or "400 timeout fetching". So I added all 8,831 URLs to the sitemap.xml and added the following headers: "Content-Type": "text/plain; charset=utf-8" "Content-Disposition": "inline" I can't tell if this helped, but 24 hours later, ChatGPT was able to fetch them properly 😊 Maybe AI assistants need the page to be indexed first, before being able to crawl it?
I made TrustMRR readable by AI agents. AI Agents often miss data when they have to scroll pages or run scripts. The goal is to make the startup marketplace easier to browse from ChatGPT, Claude, Gemini, and other AI assistants, so people can discover and acquire startups from chat. So I added public startup .md pages, llms.txt, and a limited /api/ai endpoint to give agents structured data. ChatGPT alone makes nearly as many daily requests as actual humans on TrustMRR. So I'm giving all my startups the best surfboard to ride the AI tsunami.
是什么 这条内容分享 TrustMRR 为 AI Agent 增加机器可读入口的实践:公开 startup Markdown 页面、llms.txt,以及一个受限的 /api/ai 接口。作者还将 8,831 个 URL 加入 sitemap.xml,并设置纯文本 Content-Type 与 inline Content-Disposition,以改善新页面被 AI 助手抓取的稳定性。
核心设计 核心不是重做页面视觉,而是设计一套面向 Agent 的内容交付契约:用稳定的 .md URL 承载正文,用站点地图暴露资源范围,用结构化 API 降低脚本执行和滚动解析成本。这样,ChatGPT、Claude、Gemini 等助手可以通过更简单的 HTTP 访问发现创业公司数据。
关键洞察 这体现了网站从人类浏览入口扩展为 Agent 访问入口。作者观察到上线 24 小时后抓取成功,但也明确表示无法确认究竟是哪项改动生效,因此索引先于抓取只是待验证的假设。截图中的 startup 文件列表与计数,也说明内容库存正在被显式组织和暴露。
边界与风险 这不是严格实验,成功率变化不能归因于 sitemap 或响应头。公开数据和 API 仍需处理更新一致性、速率限制、爬虫成本、敏感信息暴露及接口滥用问题;同时,llms.txt 也不应被当作所有 AI 平台必然遵循的标准。