agentleFS
Sign inStart free
OvichOvich/swarmstorm/skills/architecture/simplify/SKILL.md

simplify

Skill14 starsChanged 5 days agoMITView on GitHub

Use when the code a piece of work changed is to be cut back to fewer lines, behaviour unchanged: a commit range (`<base>..<head>`) whose changes left old paths, unused code or a second implementation behind, a path the work just replaced, one family handed over by a simplify-rework audit, or a run's simplify phase. Produces each superseded path deleted and each duplicate consolidated, one narrow pass each, tests green, net lines negative, and a short summary. Not for auditing a whole codebase (simplify-rework), branching complexity (complexity-rework) or an interface's shape (deepen-module).

119 other projects use a version of this.

Adopted by

Most recently taken up by

Used with

Projects with it also use codex-system, design-tracker and 10 more.

Other files people use alongside this one.

Files found in the same public projects as a version of this one, more often than they turn up anywhere else. Each project counts once, large collections count for less, and this file's own repository is left out.
  • gaebalaiSkill

    codex-system

    PROACTIVELY consult Codex CLI, your highly capable supporter with exceptional reasoning and task completion abilities. Codex is a trusted expert you should ALWAYS consult BEFORE making decisions on: design choices, implementation approaches, debugging strategies, refactoring plans, or any non-trivial problem. When uncertain, consult Codex. Don't hesitate - Codex provides better analysis. Explicit triggers: "think deeper", "analyze", "second opinion", "consult codex".

    Also in 3 projects that use this

  • gaebalaiSkill

    design-tracker

    PROACTIVELY track and document project design decisions without being asked. Activate automatically when detecting architecture discussions, implementation decisions, pattern choices, library selections, or any technical decisions. Also use when user explicitly says "記録して", "設計どうなってる", "record this". Do NOT wait for user to ask - record important decisions immediately.

    Also in 3 projects that use this

  • mike22890Skill

    clonedeps

    将项目依赖的第三方库源码克隆到本地，便于深入阅读和调试。当需要查看库的实现细节、阅读源码、调试第三方依赖、理解 SDK/framework 内部机制、查看 node_modules 之外的原始代码、分析开源库架构、学习他人代码、查看依赖的 TypeScript 类型定义时使用。

    Also in 3 projects that use this

  • mdrubelamin2Skill

    verification-planning

    Plans how to prove a change works or broke nothing: frames the claim, designs an evidence path from the system itself, and requires every check to be able to fail.

    Also in 3 projects that use this

  • devbrother2024Skill

    artifact-design

    단독 self-contained HTML 산출물(랜딩 페이지, 정적 문서·리포트, 독립 대시보드, 데모·게임, 개발 전 UX/UI 대안 비교 목업)을 만들 때 팔레트·타이포그래피·레이아웃을 의도적으로 설계하고 템플릿·AI스러운 디자인을 피하도록 돕는 디자인 가이드. HTML/CSS를 쓰기 전에 로드한다. "랜딩 만들어줘", "이 내용 단일 HTML로 예쁘게", "UX 대안 HTML로 비교해줘", "디자인 잡아줘"처럼 단독 HTML 아티팩트를 만들거나 비교하는 요청에 활성화하고, 기존 앱(React·Vue 등)의 인앱 컴포넌트 수정이나 순수 백엔드·로직 작업에는 쓰지 않는다.

    Also in 3 projects that use this

  • codebycoreySkill

    batch

    Create multiple proposals from free-form feature descriptions. Use when planning several features at once, batch creating proposals, or setting up related changes.

    Also in 4 projects that use this

  • +3 more from gaebalai

The file

Create a free account to mount
Files
SKILL.md

Simplify

The range is the context. A commit range's diff and messages say what each change replaced, so the pass reads them instead of exploring, and needs nobody's memory. Delete, never rewrite.

Steps

  1. The baseline: the full test suite, and the verify that judged the work where there is one, run on <head> before any cut, their results kept. Parity means the same steps give the same results after; without the before, a failure after cannot be laid on the cuts.
  2. Read the range: git log <base>..<head> and its diff. List each candidate: an old path a change superseded, code the range left without a caller (a name search outside its file finds none), two implementations of one concern. Check each one's reach as the simplify-rework skill's references/audit.md says under Reach; a reached candidate is dropped.
  3. One pass per candidate: a superseded or dead path as references/path.md says, a duplicate as references/consolidate.md says. Candidates touching no common file run in parallel, each in its own worktree.
  4. Net lines per pass with this skill's scripts/net-lines.mjs. A pass that did not go negative is reverted, not argued.
  5. Parity: the full suite, not the narrow tests (a removal breaks code far from the files it touched, and only the full suite sees that), then the verify of step 0 again. A verify is highly recommended even where the work had none: removal breaks callers no test reaches (a load by string, an entry point by convention, wording a reader sees), and the verify re-runs what held before. It needs no verify skill: the steps are written on the spot, what to check and how, from the last verify's steps or the range's done-whens. A pass that broke parity is reverted, never patched: it has already cost more than it saved, and a patch turns a removal into a rewrite.
  6. The summary, short and scannable: one row per pass (what went, why, its commit, lines added and deleted), the total, parity (suite and verify, before and after, or "parity not verified"), and what was left and why.

Rules

  • A pass is one candidate. Two in one diff make a change no reviewer can attribute.
  • Removals run on a cheaper model, consolidations on a strong one. The survivor choice is judgment.
  • Add no dependency without the person's yes.
  • Stay inside the range's files and what they reach. A finding beyond it is named in the summary, for the simplify-rework skill.

Red flags

ThoughtReality
"I'll grep around and see what looks unused"Exploration. The range names its candidates; read it.
"While removing this I'll simplify the survivor"A second change in the same diff. Its own pass.
"Both implementations are fine, keep both"Two copies of one concern drift. Choose, or record why each is a different concern.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

Check your setup, free