---
title: "什么是 AI Agents？"
canonical: https://wavect.io/zh/glossary/ai-agents/
language: zh
description: "一种使用 LLM 决定下一步、调用工具执行决定、并循环直到完成目标的软件形态。"
image: "https://wavect.io/img/general/bak/open_graph_preview.jpg"
---

技术

# AI Agents

一种使用 LLM 决定下一步、调用工具执行决定、并循环直到完成目标的软件形态。

最近审阅: 2026-05-24 审阅人 [Kevin Riedl](/zh/team/kevin-riedl/) [wiki ↗](https://www.wikidata.org/wiki/Q139796365)

AI Agent 是这套循环，不是模型本身。模型（Claude、GPT、Gemini、开放权重 LLM）是决策者。Agent 是运行时：它给模型一个目标，让它调用工具（一个数据库、一个 API、一个代码解释器、另一个 Agent），常常经由 [MCP](/zh/glossary/mcp/) ，观察结果，并把循环喂下去直到完成。

这个区分要紧，因为 Agent 式系统的失败方式与直接的 LLM 应用不同。聊天机器人答错了，用户翻篇。一个答错的 Agent 会发出一封邮件、做一笔交易，或删一份文件。生产级 Agent 需要授权 Gate、可观测性和熔断器，而多数原型都跳过了这些。

最常见的生产失败，举个实际例子：一个客服 Agent 被给了一个「解决工单」的目标和一组工具。一张畸形工单把它送进一个循环，它把同一个查询工具调了几十次，一下午就烧光了模型预算，却从未抵达一个解决方案。修法不是一个更聪明的 Prompt，而是工程：一个硬性步数上限、一个成本天花板，以及一个会停下循环并升级给人的熔断器。任何会写的动作（发一条消息、做一笔支付、删一条记录）都加一个 Human-in-the-Loop 的 Gate；只读的循环通常可以无人值守地跑。

诚实的取舍与该避免的创始人错误：Agent 增加了非确定性、延迟，以及比固定流水线大得多的爆炸半径，换来的是处理那些步骤无法预先枚举的任务。多数被吹成「Agentic」的工作流其实是被充分理解的，用一条确定性流水线、中间放一两次 LLM 调用、常常再由 [RAG](/zh/glossary/rag/) 提供接地，会服务得更好。Wavect 的立场：先问 AI 是不是这里合适的工具，再问 Agent 是不是合适的 AI 形态。如果你能把步骤写下来，就别让模型即兴去演它们。

// FAQ

## 常见问题

### 什么时候 Agent 是错的形态？

只要一条确定性流水线、中间放一两次 LLM 调用就能完成这份工作，就别上 Agent。Agent 在失火时会增加非确定性、延迟，以及大得多的爆炸半径。如果你的工作流是被充分理解的，就别让模型即兴去演那些步骤。

### 生产 Agent 最常见的失败是什么？

无界的工具调用。Agent 陷入循环、把同一个工具调了十次、烧光预算，却从未抵达目标。生产级 Agent 需要步数上限、成本天花板和熔断器。没有这些，一个糟糕的 Prompt 就会变成一次计费事故。

### Agent 需要 Human-in-the-Loop 吗？

对任何破坏性或不可逆的动作：需要。发一封邮件、做一笔支付、删一条记录、公开发帖、调用一个有副作用的外部 API。只读的循环通常可以无人值守地跑；写一侧的动作应当加一道 Gate。

相关术语

- [MCP](/zh/glossary/mcp/)
- [RAG](/zh/glossary/rag/)

相关服务

- [人工智能](/zh/services/artificial-intelligence/)

相关案例研究

- [PromptID](/zh/case-studies/promptid/)
- [Twinsoft AI](/zh/case-studies/twinsoft-ai/)
- [Quivr](/zh/case-studies/quivr/)
- [Hyperstate AI](/zh/case-studies/hyperstate-ai/)

相关指南

- [如何为 MVP 选择技术栈](/zh/software-development-guide/how-to-choose-a-tech-stack-for-mvp/)

相关博客文章

- [为什么 AI 代理项目会被砍掉](/zh/blog/why-ai-agent-projects-get-cancelled/)
- [编排 AI 代理时，专注力才是瓶颈](/zh/blog/focus-bottleneck-orchestrating-ai-agents/)

聊聊

## Structured Data

```json
{
  "@context": "https://schema.org",
  "@graph": [
    {
      "@id": "https://wavect.io/#organization",
      "@type": [
        "Organization",
        "ProfessionalService",
        "LocalBusiness"
      ],
      "employee": [
        {
          "@id": "https://wavect.io/team/kevin-riedl/#person",
          "@type": "Person",
          "jobTitle": "Managing Director",
          "name": "Kevin Riedl",
          "url": "https://wavect.io/team/kevin-riedl/",
          "worksFor": {
            "@id": "https://wavect.io/#organization",
            "@type": [
              "Organization",
              "ProfessionalService",
              "LocalBusiness"
            ]
          }
        },
        {
          "@id": "https://wavect.io/team/christof-jori/#person",
          "@type": "Person",
          "jobTitle": "Managing Director",
          "name": "Christof Jori",
          "url": "https://wavect.io/team/christof-jori/",
          "worksFor": {
            "@id": "https://wavect.io/#organization",
            "@type": [
              "Organization",
              "ProfessionalService",
              "LocalBusiness"
            ]
          }
        }
      ],
      "founder": [
        {
          "@id": "https://wavect.io/team/kevin-riedl/#person",
          "@type": "Person",
          "jobTitle": "Managing Director",
          "name": "Kevin Riedl",
          "url": "https://wavect.io/team/kevin-riedl/",
          "worksFor": {
            "@id": "https://wavect.io/#organization",
            "@type": [
              "Organization",
              "ProfessionalService",
              "LocalBusiness"
            ]
          }
        },
        {
          "@id": "https://wavect.io/team/christof-jori/#person",
          "@type": "Person",
          "jobTitle": "Managing Director",
          "name": "Christof Jori",
          "url": "https://wavect.io/team/christof-jori/",
          "worksFor": {
            "@id": "https://wavect.io/#organization",
            "@type": [
              "Organization",
              "ProfessionalService",
              "LocalBusiness"
            ]
          }
        }
      ],
      "legalRepresentative": [
        {
          "@id": "https://wavect.io/team/kevin-riedl/#person",
          "@type": "Person",
          "jobTitle": "Managing Director",
          "name": "Kevin Riedl",
          "url": "https://wavect.io/team/kevin-riedl/",
          "worksFor": {
            "@id": "https://wavect.io/#organization",
            "@type": [
              "Organization",
              "ProfessionalService",
              "LocalBusiness"
            ]
          }
        },
        {
          "@id": "https://wavect.io/team/christof-jori/#person",
          "@type": "Person",
          "jobTitle": "Managing Director",
          "name": "Christof Jori",
          "url": "https://wavect.io/team/christof-jori/",
          "worksFor": {
            "@id": "https://wavect.io/#organization",
            "@type": [
              "Organization",
              "ProfessionalService",
              "LocalBusiness"
            ]
          }
        }
      ],
      "name": "Wavect GmbH",
      "subjectOf": {
        "@id": "https://wavect.io/verified-claims.json#dataset",
        "@type": "Dataset",
        "creator": {
          "@id": "https://wavect.io/#organization",
          "@type": [
            "Organization",
            "ProfessionalService",
            "LocalBusiness"
          ]
        },
        "description": "A machine-readable registry of quantitative and qualitative claims published by Wavect, with review dates, localized page appearances and public third-party citations where available.",
        "inLanguage": "en",
        "isAccessibleForFree": true,
        "license": "https://creativecommons.org/licenses/by/4.0/",
        "name": "Wavect verified publication claims",
        "url": "https://wavect.io/verified-claims.json"
      },
      "url": "https://wavect.io/"
    },
    {
      "@id": "https://wavect.io/team/kevin-riedl/#person",
      "@type": "Person",
      "jobTitle": "Managing Director",
      "name": "Kevin Riedl",
      "sameAs": [
        "https://www.wikidata.org/wiki/Q139796365",
        "https://www.linkedin.com/in/wsdt",
        "https://github.com/wsdt"
      ],
      "url": "https://wavect.io/team/kevin-riedl/",
      "worksFor": {
        "@id": "https://wavect.io/#organization",
        "@type": [
          "Organization",
          "ProfessionalService",
          "LocalBusiness"
        ]
      }
    },
    {
      "@id": "https://wavect.io/team/christof-jori/#person",
      "@type": "Person",
      "jobTitle": "Managing Director",
      "name": "Christof Jori",
      "sameAs": [
        "https://www.wikidata.org/wiki/Q139796367",
        "https://www.linkedin.com/in/jocr77/",
        "https://github.com/jo-chris"
      ],
      "url": "https://wavect.io/team/christof-jori/",
      "worksFor": {
        "@id": "https://wavect.io/#organization",
        "@type": [
          "Organization",
          "ProfessionalService",
          "LocalBusiness"
        ]
      }
    },
    {
      "@id": "https://wavect.io/#website",
      "@type": "WebSite",
      "inLanguage": [
        "en",
        "de",
        "es",
        "zh"
      ],
      "name": "Wavect",
      "potentialAction": {
        "@type": "SearchAction",
        "query-input": "required name=search_term_string",
        "target": {
          "@type": "EntryPoint",
          "urlTemplate": "https://wavect.io/search/?q={search_term_string}"
        }
      },
      "publisher": {
        "@id": "https://wavect.io/#organization",
        "@type": [
          "Organization",
          "ProfessionalService",
          "LocalBusiness"
        ]
      },
      "url": "https://wavect.io/"
    }
  ]
}
```

```json
{
  "@context": "https://schema.org",
  "@graph": [
    {
      "@id": "https://wavect.io/zh/glossary/ai-agents/#term",
      "@type": "DefinedTerm",
      "alternateName": [
        "AI 智能体",
        "Agentic AI"
      ],
      "description": "一种使用 LLM 决定下一步、调用工具执行决定、并循环直到完成目标的软件形态。",
      "inDefinedTermSet": "https://wavect.io/zh/glossary/#termset",
      "name": "AI Agents",
      "termCode": "ai-agents",
      "url": "https://wavect.io/zh/glossary/ai-agents/"
    },
    {
      "@id": "https://wavect.io/zh/glossary/ai-agents/#webpage",
      "@type": "WebPage",
      "about": {
        "@id": "https://wavect.io/#organization",
        "@type": [
          "Organization",
          "ProfessionalService",
          "LocalBusiness"
        ]
      },
      "author": {
        "@id": "https://wavect.io/team/kevin-riedl/#person",
        "@type": "Person",
        "name": "Kevin Riedl",
        "url": "https://wavect.io/team/kevin-riedl/"
      },
      "dateModified": "2026-05-24",
      "description": "一种使用 LLM 决定下一步、调用工具执行决定、并循环直到完成目标的软件形态。",
      "headline": "AI Agents",
      "inLanguage": "zh",
      "isPartOf": {
        "@id": "https://wavect.io/#website",
        "@type": "WebSite"
      },
      "lastReviewed": "2026-05-24",
      "mainEntity": {
        "@id": "https://wavect.io/zh/glossary/ai-agents/#term"
      },
      "mentions": [
        {
          "@id": "https://wavect.io/#organization",
          "@type": [
            "Organization",
            "ProfessionalService",
            "LocalBusiness"
          ]
        },
        {
          "@type": "Service",
          "name": "人工智能",
          "url": "https://wavect.io/zh/services/artificial-intelligence/"
        }
      ],
      "name": "AI Agents",
      "reviewedBy": {
        "@id": "https://wavect.io/team/christof-jori/#person",
        "@type": "Person",
        "name": "Christof Jori",
        "url": "https://wavect.io/team/christof-jori/"
      },
      "speakable": {
        "@type": "SpeakableSpecification",
        "cssSelector": [
          ".gls-single__h1",
          ".gls-single__tldr"
        ]
      },
      "url": "https://wavect.io/zh/glossary/ai-agents/"
    },
    {
      "@type": "BreadcrumbList",
      "itemListElement": [
        {
          "@type": "ListItem",
          "item": "https://wavect.io/zh/",
          "name": "首页",
          "position": 1
        },
        {
          "@type": "ListItem",
          "item": "https://wavect.io/zh/glossary/",
          "name": "技术、产品与合作模式术语表",
          "position": 2
        },
        {
          "@type": "ListItem",
          "item": "https://wavect.io/zh/glossary/ai-agents/",
          "name": "AI Agents",
          "position": 3
        }
      ]
    }
  ]
}
```

```json
{
  "@context": "https://schema.org",
  "@type": "FAQPage",
  "mainEntity": [
    {
      "@type": "Question",
      "acceptedAnswer": {
        "@type": "Answer",
        "text": "只要一条确定性流水线、中间放一两次 LLM 调用就能完成这份工作，就别上 Agent。Agent 在失火时会增加非确定性、延迟，以及大得多的爆炸半径。如果你的工作流是被充分理解的，就别让模型即兴去演那些步骤。"
      },
      "name": "什么时候 Agent 是错的形态？"
    },
    {
      "@type": "Question",
      "acceptedAnswer": {
        "@type": "Answer",
        "text": "无界的工具调用。Agent 陷入循环、把同一个工具调了十次、烧光预算，却从未抵达目标。生产级 Agent 需要步数上限、成本天花板和熔断器。没有这些，一个糟糕的 Prompt 就会变成一次计费事故。"
      },
      "name": "生产 Agent 最常见的失败是什么？"
    },
    {
      "@type": "Question",
      "acceptedAnswer": {
        "@type": "Answer",
        "text": "对任何破坏性或不可逆的动作：需要。发一封邮件、做一笔支付、删一条记录、公开发帖、调用一个有副作用的外部 API。只读的循环通常可以无人值守地跑；写一侧的动作应当加一道 Gate。"
      },
      "name": "Agent 需要 Human-in-the-Loop 吗？"
    }
  ],
  "speakable": {
    "@type": "SpeakableSpecification",
    "cssSelector": [
      ".faq-question",
      ".faq-answer"
    ]
  }
}
```