高效商务代理剖析指南

The architecture, latency & cost techniques, and eval practices for agents that make it easier to buy and sell online.

让在线买卖更便捷的智能体架构、延迟与成本优化技术，以及评估实践。

Over the past year, we've worked with teams across the commerce industry — retailers, marketplaces, travel, entertainment, and telecom providers — to build commerce agents using Claude.

在过去的一年里，我们与电商行业的各个团队——包括零售商、市场平台、旅游、娱乐和电信提供商——合作，使用 Claude 构建电商智能体。

These agents are in production, and enterprise customers have seen larger carts and more efficient seller operations when using them. They also share a simple architecture: Claude in an agent loop equipped with a set of skills, tools, and a strong eval suite.

这些智能体已投入生产，企业客户在使用它们后看到了更大的购物车和更高效的商家运营。它们还共享一种简单的架构：处于智能体循环中的 Claude，配备了一系列技能、工具和一个强大的评估套件。

This post is for the engineers and engineering leaders building these (or other consumer facing) agents. Part 1 covers the architecture, which you decide once. Part 2 covers latency and cost. Part 3 covers production: memory, safety, evals, and scaling the work across an organization.

本文面向构建这些（或其他面向消费者）代理的工程师和工程领导者。第一部分涵盖架构，这是一次性决策。第二部分涵盖延迟和成本。第三部分涵盖生产环境：内存、安全性、评估，以及在组织内规模化推进工作。

In this guide

在本指南中

1. Part 1: The architecture
   1. What is a commerce agent?
   2. Skills, not subagents
   3. System prompt or skill: decide by frequency
   4. Engineering agent tooling
   5. The UI components are tools
2. 第一部分：架构
   1. 什么是商务代理？
   2. 技能，而非子代理
   3. 系统提示词或技能：根据频率决定
   4. 工程代理工具
   5. UI 组件即工具
3. Part 2: Making it fast and affordable
   1. Minimizing task completion latency
   2. Perceived latency
   3. Prompt caching
   4. Choosing the model and its configuration
4. 第二部分：使其快速且经济实惠
   1. 最小化任务完成延迟
   2. 感知延迟
   3. Prompt缓存
   4. 选择模型及其配置
5. Part 3: Running it in production
   1. Memory that survives the session
   2. Safety: enforcement lives in the harness
   3. Evals: shipping a non-deterministic system
   4. Shipping with a large organization
6. 第三部分：在生产环境中运行
   1. 在会话中保留的记忆
   2. 安全性：执行机制存在于测试框架中
   3. 评估：交付非确定性系统
   4. 在大型组织中交付
7. Looking ahead
8. 展望未来

01

01

The architecture

架构

One model in a standard agent loop, with skills for the long tail and tools that call the systems you already run. You decide this once.

在标准智能体循环中使用一个模型，具备处理长...