为什么AI生成的代码很容易，但工程信任却很难建立

Imagine an AI coding agent generating a large pull request filled with AI-generated code. Everything looks exactly as it should:

想象一个 AI 编码智能体生成了一个巨大的 pull request，里面充满了 AI 生成的代码。一切看起来都完全正确：

The build passes.
构建通过。
Every automated test is green.
每个自动化测试都通过。
Code review finds no major issues.
代码审查未发现重大问题。
The change is merged.
变更已合并。

Later, your team discovers the agent quietly misinterpreted an ambiguous requirement near the beginning of the implementation. Every decision that followed was internally consistent. The tests validated the implementation. The reviewer approved the diff. The code did exactly what the agent believed the requirement asked for.

后来，你的团队发现智能体在实现初期悄悄误解了一个模糊的需求。随后做出的每一个决策在内部都是自洽的。测试验证了该实现。审查者批准了代码差异。代码完全按照智能体所理解的需求要求去执行。

The AI made a mistake, but the truly frightening part is that nothing in the engineering process ever proved the agent’s interpretation was correct before the code reached production. If passing tests, code review, and a successful build still cannot prove that AI-generated code is trustworthy, what can?

AI 犯了一个错误，但真正可怕的是，在代码到达生产环境之前，工程流程中没有任何东西能证明智能体的解释是正确的。如果通过的测试、code review 和成功的构建仍然无法证明 AI 生成的代码是可信的，那还有什么可以证明呢？

At Salesforce, we encountered this challenge while using AI coding agents to develop an AI-powered mobile application designer. The agents could generate code remarkably well. Our challenge was determining why we should trust that AI-generated code.

在Salesforce，我们在使用AI编程智能体开发AI驱动的移动应用设计器时遇到了这一挑战。这些智能体能够非常出色地生成代码。我们的挑战在于确定为什么我们应该信任那些AI生成的代码。

Where Was the Mistake Supposed to Be Caught?

错误本该在哪里被捕获？

A coding agent can often find a path from a requirement to a functioning implementation. The problem is that many paths may produce the same visible result. An agent might introduce a utility even though one already exists, follow a framework convention while violating a repository-specific convention, or interpret an ambiguous requirement without revealing that it made a decision. The code can compile and still increase technical debt.

编码代理通常能找到一条从需求到可运行实现的路径。问题在于，许多路径可能会产生相...