---
id: 20261117-T0-17
title: "Vercel：Agent也是用户，你的软件需要评估体系"
title_en: "Vercel: Agents Are Users, Your Software Needs Evals"
url: https://ai.daily.yangsir.net/daily/20261117-T0-17
issue_date: 2026-11-17
publish_date: 2026-11-16T05:00:00.000Z
category: insight
source_name: "Vercel Blog"
source_url: https://vercel.com/blog/agent-evals
---

# Vercel：Agent也是用户，你的软件需要评估体系

Vercel发文指出，Agent正在以人类代理人的身份使用软件，无论是开发Agent的一方，还是提供API、CLI、MCP服务器、SDK等被Agent调用的软件方，都需要通过评估（evals）来衡量任务完成质量。文章强调，Agent不会看你的产品发布视频或高速公路广告牌，它们只按接口和文档行事。对开发者而言，这意味着产品体验的评判标准要从"人好不好用"扩展到"Agent能不能可靠跑通"，否则Agent在调用你的服务时出错，你甚至不会收到反馈。

## English Version

**Vercel: Agents Are Users, Your Software Needs Evals**

Vercel argues that agents have become users of software, acting on people's behalf, and that evals are the way to measure how well delegated work gets done. Whether you build the agent or the software it calls—API, CLI, MCP server, SDK—you need a way to quantify task success. The post notes agents never saw your launch video or your billboard on Highway 101; they interact purely through interfaces and docs. For developers, this shifts product quality from human usability to agent reliability, and failures may go unnoticed without systematic evaluation.

---

**来源**：[Vercel Blog](https://vercel.com/blog/agent-evals)

**详情页**：https://ai.daily.yangsir.net/daily/20261117-T0-17

---

*智语观潮 · Daily — https://ai.daily.yangsir.net/llms.txt*