AI Agent生产环境仍存在五类工程缺陷

大多数 AI 工具只会告诉你它们擅长什么。这样你只会在生产环境中才发现那些差距。所以这里有另一半：我们的 AI Agent 具体还在哪些地方会失误，以及我们对每一个问题正在做什么。

What it's genuinely good at

真正擅长的方面

Targeted bug-fixing (~85–90%) — given real broken code, it produces a fix that compiles clean and makes the specific change.

定向 bug 修复（约 85–90%）—— 给定真正有问题的代码，它能生成一个编译通过且做出了具体修改的修复方案。

Simple full-stack builds — it scaffolds working SPAs reliably.

简单的全栈构建——它能可靠地搭建可用的 SPA。

Prompt → full-stack app → live deploy with a real database — validated end to end.

Prompt → 全栈应用 → 使用真实数据库的在线部署——端到端验证。

Where it still misses

仍然不足的地方

From our own correctness-gated evals, the honest gaps:

根据我们的正确性门控评估，真实的差距如下：

Complex full-stack scaffolding. On a notes app with auth, it built the database schema but missed declaring the bcrypt / jsonwebtoken dependencies — the kind of gap that compiles in isolation but breaks the build.

复杂全栈搭建。在一个带认证的笔记应用中，它构建了数据库架构但遗漏了 bcrypt / jsonwebtoken 依赖声明——这类差距在独立编译时能通过，但会破坏整体构建。

Run-to-run consistency. A landing-page fixture's dev-server check passed on one run and failed on another. Non-determinism is real, and we measure it instead of hiding it.

运行间的一致性。着陆页 fixture 的开发服务器检查在一次运行中通过，但在另一次运行中失败。非确定性是真实存在的，我们对其进行测量而不是隐藏它。

Small correctness slips. Using React.Fragment without importing it; a to-do app that skipped localStorage persistence. Individually minor — exactly the sort of thing you'd want to catch in a diff review, not in production.

小的正确性失误。使用 React.Fragment 但没有导入它；待办事项应用跳过了 localStorage 持久化。每个单独看都很小——正是那种你希望在 diff 审查中捕获而不是在生产环境中才发现的问题。

Read-only tasks. On a task that only asked it to find and report something, it wrote files instead of answering. Knowing when not to act is its own skill.

只读任务。在一个只要求它查找并报告某些内容的任务中，它却写了文件而不是回答。知道何时不行动本身就是一种技能。

Every genuine failure feeds a closed self-improvement loop: the run is analyzed, a prompt change is proposed, and the candidate is validated against the same correctness gate — with a holdout set guarding against regressions — before anything reaches production. Some of the misses above already have fixes in flight; some don't yet. We publish both states.

每个真正的失败都会进入一个闭环自我改进循环：分析运行结果，提出 prompt 变更建议，在相同的正确性门控下验证候选方案——用保留集防止回归——然后才会进入生产环境。上述的一些失误已经在修复中；一些还没有。我们会发布两种状态。

Why publish this at all

为什么要发布这些

Because you're going to hit these limits either way — the only question is whether you find out from us now or from a broken build later. It's also why NoCoder puts diff review at the center: the agent isn't perfect, so you stay in the loop. Every change is a per-file diff you approve before it applies.

因为无论如何你都会遇到这些限制——唯一的问题是你是现在从我们这里得知，还是稍后从一次构建失败中得知。这也是 NoCoder 将 diff 审查置于核心的原因：Agent 不是完美的，所以你需要保持在循环中。每个变更都是逐文件的差分，在应用之前由你批准。

The short version: Our agent is strong at bug-fixing (~85–90%) and simple full-stack builds, and weaker on complex scaffolding, consistency, and a few specific tasks (missed auth deps, a skipped localStorage step, a Fragment import, a read-only find task). We fix what we can via a validated self-improvement loop and publish the rest. See how we benchmark or try NoCoder free.

简而言之：我们的 Agent 在 bug 修复（约 85–90%）和简单全栈构建方面表现强劲，在复杂脚手架、一致性和一些特定任务（遗漏认证依赖、跳过 localStorage 步骤、Fragment 导入、只读查找任务）方面表现较弱。我们通过经过验证的自我改进循环修复我们能修复的，并发布其余的。

This post originally appeared on the NoCoder blog. NoCoder is an AI full-stack app builder that shows every change as a diff you approve before it ships — build at AI speed, ship only what you've reviewed. Free to start on free models.

这篇文章最初发表在 NoCoder 博客上。NoCoder 是一个 AI 全栈应用构建器，在发布之前将每个变更展示为你批准的差分——以 AI 速度构建，只发布你审查过的内容。免费开始使用免费模型。