I’ve always thought about AI agents as microservices+.我一直把AI智能体看作是微服务+。
They need everything a traditional microservice needs, and:它们需要传统微服务所需的一切，此外：
- More authentication requirements because an agent may act on behalf of many different users.更多的身份验证要求，因为智能体可能代表许多不同的用户行事。
- More policy requirements because an agent’s behavior can be less predictable, requiring guardrails and policy enforcement.更多的策略要求，因为智能体的行为可能不太可预测，需要护栏和策略执行。
- More observability requirements, especially around context, prompts, tool calls, and the contents of requests and responses.更多的可观测性要求，尤其是在上下文、提示、工具调用以及请求和响应的内容方面。
When thinking about agent auth, I found myself reflecting on a traffic lawyer I hired years ago after receiving a traffic ticket for failing to stop for a school bus. It was my first, and so far only, traffic ticket.😅在思考智能体身份验证时，我回想起几年前因未给校车停车而收到交通罚单后聘请的一位交通律师。那是我第一次，也是迄今为止唯一一次交通罚单。😅
The experience turned out to be a useful mental model for understanding agent auth.这段经历后来成为理解智能体身份验证的有用思维模型。
Imagine a lawyer walking into court to represent Alice.想象一位律师走进法庭代表爱丽丝。
This is similar to an AI agent receiving a request from Alice and performing actions on her behalf.这类似于AI智能体收到爱丽丝的请求并代表她执行操作。
The judge first asks the lawyer to prove who he is.法官首先要求律师证明自己的身份。
This is agent identity. Before the system can trust an agent, it needs to know exactly which agent is making the request.这就是智能体身份。在系统信任智能体之前，它需要确切知道是哪个智能体在发出请求。
Next, the judge asks, “Who are you representing today?”接下来，法官问：“你今天代表谁？”
This is principal identity. The system needs to know not only who the agent is, but also which user the agent is acting for.这就是主体身份。系统不仅需要知道智能体是谁，还需要知道智能体代表哪个用户。
The lawyer then presents documentation showing that he is authorized to represent Alice in this specific case.然后律师出示文件，证明他有权在此特定案件中代表爱丽丝。
In agent systems, this is often represented by an On-Behalf-Of (OBO) token or another delegation artifact. The token carries information about:在智能体系统中，这通常由代表令牌（OBO令牌）或其他委托工件表示。该令牌携带以下信息：
- The identity of the principal (Alice)主体（爱丽丝）的身份
- The identity of the agent智能体的身份
- The delegated permissions被委托的权限
- The scope of the delegation委托的范围
At this point, the judge knows three things:此时，法官知道三件事：
- Who the lawyer is律师是谁
- Who the lawyer represents律师代表谁
- What authority has been delegated to the lawyer律师被授予了什么权限
But that still isn’t enough.但这仍然不够。
The judge must also verify that the lawyer is allowed to represent Alice in this particular traffic case. This is where policy enforcement comes in.法官还必须核实律师是否被允许在此特定交通案件中代表爱丽丝。这就是策略执行发挥作用的地方。
Having a valid delegation does not automatically grant unlimited access. The requested action must still comply with the applicable policies and scopes.拥有有效的委托并不自动授予无限访问权限。请求的操作仍必须符合适用的策略和范围。
In a real courtroom, the lawyer and the judge handle most of this complexity. They carry identities, verify credentials, validate representation rights, and enforce the rules of the court.在真实的法庭中，律师和法官处理了大部分复杂性。他们携带身份、验证凭证、确认代表权并执行法庭规则。
In an agentic system, we need similar infrastructure.在智能体系统中，我们需要类似的基础设施。
An agent platform must be able to:智能体平台必须能够：
- Establish strong agent identities建立强大的智能体身份
- Carry principal identities across requests在请求中携带主体身份
- Issue and validate delegation tokens签发和验证委托令牌
- Enforce authorization policies and scopes执行授权策略和范围
- Provide observability and audit trails for agent actions为智能体操作提供可观测性和审计追踪
This is where an AI native gateway can play an important role.这就是AI原生网关可以发挥重要作用的地方。
Rather than requiring every agent to independently implement identity propagation, delegation verification, policy enforcement, and auditing, the agent gateway and mesh can centralize these capabilities. The agent gateway and mesh become the equivalent of the court clerk, bailiff, and records office combined: ensuring identities are verified, delegations are valid, policies are enforced, and actions are auditable.与其要求每个智能体独立实现身份传播、委托验证、策略执行和审计，智能体网关和网格可以集中这些能力。智能体网关和网格相当于法庭书记员、法警和档案室的结合：确保身份得到验证、委托有效、策略得到执行、操作可审计。
Combined with existing identity and service-mesh technologies such as SPIFFE, cert-manager, Istio, and agentgateway, we can build an agent platform where agents focus on business logic while the platform handles identity, delegation, policy enforcement, and observability.结合现有的身份和服务网格技术，如SPIFFE、cert-manager、Istio和agentgateway，我们可以构建一个智能体平台，让智能体专注于业务逻辑，而平台处理身份、委托、策略执行和可观测性。
The core idea is simple:核心思想很简单：
A lawyer is not the client.律师不是客户。
An agent is not the user.智能体不是用户。
Both operate with their own identities while acting on behalf of someone else, under a specific delegation and within a defined scope. Agent auth is fundamentally about making that relationship explicit, verifiable, and enforceable.两者都以自己的身份运作，同时代表他人，在特定的委托和定义的范围下行事。智能体身份验证从根本上讲就是使这种关系明确、可验证和可执行。