基于 SOC 职业分类
用 Codex 或 Claude 帮你安装 复制这段 Prompt,粘贴到 Codex、Claude 或其他助手里,让它检查 Skill 页面并帮你完成安装。
直接命令不会经过审查 Prompt;运行前请先检查来源。
npx skills add https://github.com/vignesh2027/AI-AGENT-SKILLS --skill code-simplification命令会保持在同一行。复制前请横向滚动并检查完整内容。
想先保存到本地?可下载 SkillsMP 当前能够提供的文件。
正在显示 SKILL.md
| name | code-simplification |
| description | Remove accidental complexity without changing behavior — apply Chesterton's Fence |
| difficulty | senior |
| domains | ["general"] |
Every line of code is a liability. Complexity that serves no purpose makes systems harder to understand, harder to test, and harder to change. This skill removes accidental complexity while preserving intentional complexity and essential behavior.
/simplify workflowBefore removing any code that seems unnecessary: understand why it was written. If you don't know why a fence exists, don't remove it. Find out first.
Only remove accidental complexity. Essential complexity cannot be removed — only managed.
Dead code: Code that is never executed. Remove it (git remembers it).
Premature abstraction: Abstractions that are more complex than the thing they abstract. Inline them.
Unnecessary indirection: Function A calls function B which calls function C and each is 5 lines. Collapse them.
Redundant comments: Comments that say what the code already says. Delete them. Keep comments that explain WHY.
Magic numbers: Replace if (retries > 3) with if (retries > MAX_RETRIES).
Over-engineered error handling: 10 levels of nested try/catch for a case that can't happen. Simplify.
Duplicate logic: The same logic in 3 places. Extract to one place.
One simplification at a time. Run tests after each step.
The test suite must pass. If there are no tests, write tests for the behavior you're simplifying before you simplify it.
Have someone unfamiliar with this code read it. Is it easier to understand than before?
"Removing this might break something I don't know about" This is Chesterton's Fence. Understand it first. Then remove it.
"The abstraction makes it more extensible" If you are not extending it now, the abstraction is premature. Three identical lines are better than a wrong abstraction.
Build UIs that work for all users including keyboard navigation, screen readers, and WCAG 2.2
Design multi-agent systems with robust tool interfaces, state management, and failure handling
Build ML systems with disciplined training, evaluation, deployment, and safety practices