T
traeai
登录
返回首页
David Heinemeier Hansson

Basecamp becomes agent accessible

8.5Score
Basecamp becomes agent accessible
AI 深度提炼
  • SaaS产品应从“AI功能内嵌”转向“Agent可访问”架构,通过标准化API与CLI赋予智能体完整操作权限。
  • MCP协议与工具调用能力正成为AI集成新标准,可替代传统繁琐的API对接,实现跨系统自动化工作流。
  • 智能体将依托ChatGPT等主流入口快速普及,企业需提前布局Agent友好的接口设计以抢占生态先机。
#AI Agent#SaaS架构#MCP协议#API设计#37signals
打开原文

![Image 1 David Heinemeier Hansson](https://world.hey.com/dhh) March 25, 2026

In the past 18 months, we've experimented with a ton of AI-infused features at 37signals. Fizzy had all sorts of attempts. As did Basecamp. But as Microsoft and many others have realized, it's not that easy to make something that's actually _good_ and would welcomed by users. So we didn't ship.

In the meantime, agents have emerged has the killer app for AI. Not only are LLMs much smarter when they can check their thinking using tools, but the file system also gives them the memory implant they needed to learn between prompts. And now they can actually do stuff!

Not only can you have your agent look through everything in Basecamp, summarize whatever you need, but it can also set up to-do lists, post message updates, chat with humans and clankers alike, upload reference files, and arrange a project schedule. Anything you can do in Basecamp, agents can now do too.

This becomes extra powerful when you combine Basecamp with all the other tools you might be using that are also agent accessible. For software development, you can use the MCP from Sentry to trawl through major sources of bugs, then have the agent summarize that in a message for Basecamp. Or you have it download, analyze, and highlight key customer complaints by giving it access to your help desk system.

All this was possible in the past with APIs, hand-written integrations, and human data scientists. But it was cumbersome, slow, and expensive, so most people just didn't. A vanishingly small portion of Basecamp customers have ever directly interacted with our API. But agents? I think adoption is going to be swift.

Not because everyone is going to run OpenCode, Claude Code, or Gemini CLI. But because agents are going to be incorporated into ChatGPT, Gemini, Grok, and all the other mainstream interfaces who were collectively embarrassed by OpenClaw's meteoric ascent and popularity very quickly. There's a huge demand out there for a personal agent that can act as your private executive assistant.

This is where the puck is going, and we're skating to meet it with agent accessibility across the board. Basecamp is first, Fizzy is next, and we'll hit HEY before long too. Revamped APIs, comprehensive CLIs, and the skills to use them whatever your harness or claws look like.

About David Heinemeier Hansson

问问这篇内容

回答仅基于本篇材料
    0 / 500

    Skill 包

    领域模板,一键产出结构化笔记
    • 论文精读包

      把一篇论文 / 技术博客精读成结构化笔记:问题、方法、实验、批判、延伸阅读。

      • · TL;DR(1 段)
      • · 研究问题与动机
      • · 方法概览
    • 投融资雷达包

      把一条融资 / 创投新闻整理成投资人视角的雷达卡:交易要点、判断、竞争格局、风险、尽调清单。

      • · 交易要点(公司 / 轮次 / 金额 / 投资人 / 估值,材料未明示则写 “未披露”)
      • · 投资 thesis(这家公司为什么值得关注)
      • · 竞争格局与替代方案

    导出到第二大脑

    支持 Notion / Obsidian / Readwise
    下载 Markdown(Obsidian 直接拖入)