<?xml version="1.0" encoding="utf-8" standalone="yes"?><rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom"><channel><title>Capybara Lab / 水豚数智社 on Capybara Lab</title><link>https://www.capybara-lab.com/</link><description>Recent content in Capybara Lab / 水豚数智社 on Capybara Lab</description><generator>Hugo -- gohugo.io</generator><language>en-us</language><lastBuildDate>Sun, 05 Apr 2026 00:00:00 +0000</lastBuildDate><atom:link href="https://www.capybara-lab.com/index.xml" rel="self" type="application/rss+xml"/><item><title>告别高昂API账单：Serverless GPU托管大模型实战指南（深度解析版）</title><link>https://www.capybara-lab.com/articles/serverless-gpu-hosting-llm-cost-guide/</link><pubDate>Sun, 05 Apr 2026 00:00:00 +0000</pubDate><guid>https://www.capybara-lab.com/articles/serverless-gpu-hosting-llm-cost-guide/</guid><description>拆解商业 API 的隐性成本，并对比 Modal、Baseten、RunPod、ZeroGPU 等方案，理解如何用 Serverless GPU 低成本托管自己的大模型。</description></item><item><title>数智水豚AI周报（2026/04/05）：当资本狂欢遇上应用迷局</title><link>https://www.capybara-lab.com/articles/weekly-updates-2026-04-05/</link><pubDate>Sun, 05 Apr 2026 00:00:00 +0000</pubDate><guid>https://www.capybara-lab.com/articles/weekly-updates-2026-04-05/</guid><description>从融资狂潮、模型进展到应用落地与安全焦虑，梳理 2026 年 4 月 5 日这一周 AI 世界最值得关注的 22 个切面。</description></item><item><title>告别“失忆”：为 AI Agent 注入长效记忆的架构与方法论</title><link>https://www.capybara-lab.com/articles/agent-long-term-memory-architecture/</link><pubDate>Sat, 04 Apr 2026 00:00:00 +0000</pubDate><guid>https://www.capybara-lab.com/articles/agent-long-term-memory-architecture/</guid><description>从超大上下文窗口的局限讲起，系统梳理 Agent 长效记忆的类型、生命周期、更新路径，以及它与 RAG 的关键区别。</description></item><item><title>AI本周新见闻（2026/03/29）</title><link>https://www.capybara-lab.com/articles/weekly-updates-2026-03-29/</link><pubDate>Sun, 29 Mar 2026 00:00:00 +0000</pubDate><guid>https://www.capybara-lab.com/articles/weekly-updates-2026-03-29/</guid><description>从 ERP、营销、电商到安全治理，观察 Agent 进入执行层后，企业落地为何同时踩下油门与刹车。</description></item><item><title>告别"失忆"：如何为你的 AI Agent 打造真实可用的长效记忆？</title><link>https://www.capybara-lab.com/articles/agent-long-term-memory/</link><pubDate>Sun, 29 Mar 2026 00:00:00 +0000</pubDate><guid>https://www.capybara-lab.com/articles/agent-long-term-memory/</guid><description>从上下文窗口的局限讲起，系统拆解 Agent 长效记忆的核心类型、生命周期，以及热路径与后台更新的工程权衡。</description></item><item><title>AI 智能体终迎通用语：深度解读 A2A 协议，开启协作新纪元</title><link>https://www.capybara-lab.com/articles/a2a-protocol-multi-agent-collaboration/</link><pubDate>Sun, 22 Mar 2026 00:00:00 +0000</pubDate><guid>https://www.capybara-lab.com/articles/a2a-protocol-multi-agent-collaboration/</guid><description>从 A2A 协议的设计目标、角色分工与交互方式出发，理解多智能体协作为什么开始走向标准化。</description></item><item><title>AI本周新见闻（2026/03/22）</title><link>https://www.capybara-lab.com/articles/weekly-updates-2026-03-22/</link><pubDate>Sun, 22 Mar 2026 00:00:00 +0000</pubDate><guid>https://www.capybara-lab.com/articles/weekly-updates-2026-03-22/</guid><description>从端侧 Agent、算力协议到 AI FinOps 与监管动作，本周焦点集中在 AI 基础设施如何走向可持续运行。</description></item><item><title>AI本周新见闻（2026/03/15）</title><link>https://www.capybara-lab.com/articles/weekly-updates-2026-03-15/</link><pubDate>Sun, 15 Mar 2026 00:00:00 +0000</pubDate><guid>https://www.capybara-lab.com/articles/weekly-updates-2026-03-15/</guid><description>当 Agent 开始渗透企业工单、连接器与运营系统，AI 的核心命题也从能力展示转向治理、调试与可控部署。</description></item><item><title>AI本周新见闻（2026/03/08）</title><link>https://www.capybara-lab.com/articles/weekly-updates-2026-03-08/</link><pubDate>Sun, 08 Mar 2026 00:00:00 +0000</pubDate><guid>https://www.capybara-lab.com/articles/weekly-updates-2026-03-08/</guid><description>当算法效率快速突破，而企业数据与系统准备仍明显滞后，本周最值得关注的是技术加速度与组织惯性的正面碰撞。</description></item><item><title>手把手教你构建 Agent Skill：让你的 AI 从此告别“健忘症”</title><link>https://www.capybara-lab.com/articles/build-agent-skill-from-scratch-cn/</link><pubDate>Sun, 08 Mar 2026 00:00:00 +0000</pubDate><guid>https://www.capybara-lab.com/articles/build-agent-skill-from-scratch-cn/</guid><description>从判断任务是否值得封装开始，逐步讲清 Agent Skill 的结构、触发方式与实战写法。</description></item><item><title>给你的 AI 装个“技能包”，让它从“万事通”变“专家”</title><link>https://www.capybara-lab.com/articles/agent-skills-from-generalist-to-specialist/</link><pubDate>Sun, 01 Mar 2026 00:00:00 +0000</pubDate><guid>https://www.capybara-lab.com/articles/agent-skills-from-generalist-to-specialist/</guid><description>解释 Agent Skills 的核心机制，以及它为什么能让通用 AI 逐步变成可复用的领域专家。</description></item><item><title>为什么要做 Capybara Lab / 水豚数智社</title><link>https://www.capybara-lab.com/articles/first-post/</link><pubDate>Mon, 16 Feb 2026 00:00:00 +0000</pubDate><guid>https://www.capybara-lab.com/articles/first-post/</guid><description>这是 Capybara Lab / 水豚数智社的第一篇文章：解释为什么要做这个网站、这里会写什么，以及接下来值得期待什么。</description></item><item><title/><link>https://www.capybara-lab.com/uploads/articles/readme/</link><pubDate>Mon, 01 Jan 0001 00:00:00 +0000</pubDate><guid>https://www.capybara-lab.com/uploads/articles/readme/</guid><description>Article upload folder Put your raw article materials here.
Recommended structure Create one subfolder per article, for example:
content/uploads/articles/my-first-article/ content/uploads/articles/ai-market-notes-april/ Inside each article folder, upload:
article.md — the article text cover.jpg or cover.png — optional cover image any inline images used by the article Example content/uploads/articles/my-first-article/article.md content/uploads/articles/my-first-article/cover.jpg content/uploads/articles/my-first-article/chart-1.png
After you upload them, tell me which folders are ready and I’ll convert them into published site articles under content/articles/.</description></item><item><title>About</title><link>https://www.capybara-lab.com/about/</link><pubDate>Mon, 01 Jan 0001 00:00:00 +0000</pubDate><guid>https://www.capybara-lab.com/about/</guid><description>I write about AI, tools, products, and industry shifts.</description></item><item><title>Talk to Digital Capybara</title><link>https://www.capybara-lab.com/talk/</link><pubDate>Mon, 01 Jan 0001 00:00:00 +0000</pubDate><guid>https://www.capybara-lab.com/talk/</guid><description/></item></channel></rss>