闭环：如何利用自动化反馈循环构建自我改进的AI系统

After six cycles, the system ran out of things to fix. That sentence probably raises more questions than it answers. Here is how we got there.

经过六个周期后，系统已经没有什么可修复的了。这句话引发的疑问可能比解答的还要多。以下是我们是如何走到这一步的。

The first skill PR took three days to merge. Twelve comments on structure. Six on conventions. The contributor fixed everything, merged, moved on. The next contributor hit the same twelve walls.

第一个skill PR花了三天时间才合并。关于结构有十二条意见。关于约定有六条意见。贡献者修复了所有问题，合并，继续前进。下一个贡献者撞上了同样的十二堵墙。

If you have reviewed enough pull requests, you have probably seen this pattern. An expert writes a thoughtful review comment. That comment fixes one PR. Then the wisdom evaporates. The next person has no way to benefit from it. Multiply that across dozens of contributors and hundreds of PRs, and you have an organization that generates enormous amounts of insight and retains almost none of it.

如果你审查过足够多的拉取请求，你可能见过这种模式。一位专家写了一条深思熟虑的审查评论。该评论修复了一个 PR。然后这些智慧就蒸发了。下一个人无法从中受益。将这种情况乘以数十名贡献者和数百个 PR，你就得到了一个产生大量见解却几乎不保留任何见解的组织。

This is not a tooling gap. It is an architectural one. The signal already exists in the review process. Nobody is collecting it, synthesizing it, or feeding it back into the AI system that produced the artifact in the first place.

这不是工具层面的差距，而是架构层面的差距。信号已经存在于评审过程中。没有人去收集它、综合它，或者将其反馈到最初生成该产物的AI系统中。

We Built a Generator

我们构建了一个生成器

The breakthrough was not realizing we needed better reviewers, rather, it was realizing that repeated review comments were not really comments anymore, they were undocumented requirements. Once the same guidance appears often enough, it no longer belongs in a human review. It belongs in the system itself.

突破不在于意识到我们需要更好的审查者，而在于意识到重复的审查意见实际上已经不再是意见了，它们是未记录的需求。一旦相同的指导出现得足够频繁，它就不再属于人工审查的范畴。它应该属于系统本身。

So we did the obvious thing: we built a skill that creates other skills. We call it creating-sf-skill, a meta skill that takes a user’s intent and produces a complete skill directory: SKILL.md with frontmatter, reference files, the full structure. It encodes the conventions that reviewers kept repeating: gerund naming, pro...