kill-slop
Skill27 starsChanged 4 months agoMITView on GitHub
Stop AI coding agents from shipping sloppy code — over-abstraction, drive-by refactors, obvious comments, extra files, unrequested deps, giant unreadable files, or fishing outside the workspace for “inspiration”. Use when writing or editing code, implementing features, fixing bugs, splitting or organizing files, or when the user mentions slop, cleanup, keep it simple, minimal diff, file too big, hard to navigate, or stay in the repo.
2 other projects use a version of this, 1 of them edited.
Adopted by
Most recently taken up by
- rafaelob/pragmatic-agent-skillsedited, first seen Nov 1, 2026
- aixarizzo/kill-sloptheir own version, first seen Jul 31, 2026
Other versions2 projects use a version of this.
Word for word · 0
None found.
Edited · 1
- rafaelob/pragmatic-agent-skills★ 2411% same
Wrote their own version · 1
- aixarizzo/kill-slop★ 17same name
How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.
In this repository2 other files it gives its agents.
- pr-hygieneSkill
- security-reviewSkill
The file
Show its rules and text
We split this skill into the rules it gives an agent
28 rules. About 0 are common, 25 are distinctive, and 3 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈25
Distinctive · estimate
This author's own choices. Worth a look.
3
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 18
NEVER create a file unless the task clearly needs it or a navigability split below applies
- Distinctive: line 19
NEVER add a dependency unless the user asked or existing code already requires it
- Distinctive: line 20
NEVER refactor unrelated code "while you're here"
- Distinctive: line 21
NEVER add comments that restate what the code already says
- Distinctive: line 22
NEVER introduce abstractions (wrappers, helpers, factories, "utils") for a single use
- Distinctive: line 23
NEVER expand scope beyond the user's request
- Distinctive: line 24
NEVER invent TODOs, stubs, or "for later" scaffolding
- Distinctive: line 25
NEVER search, list, or read outside the workspace root unless the user explicitly asks
- Distinctive: line 26
NEVER browse sibling projects, home, Desktop, or prior chats for “patterns” or inspiration
- Very specific: line 27
NEVER invent Clean Architecture / SOLID folder trees (domain/, application/, infrastructure/, usecases/, …) unless the repo already uses …
- Distinctive: line 28
MUST match existing project patterns before inventing new ones
- Distinctive: line 29
MUST prefer editing an existing file over adding a new one
- Distinctive: line 30
MUST delete dead code you introduce.
- Distinctive: line 34
Default to minimal. If the user explicitly asks for more structure, obey that request.
- Distinctive: line 44
Do not override an explicit ask in the name of kill-slop.
- Distinctive: line 58
If only (2) is true (“I could extract this”), do not split.
- Distinctive: line 62
MUST put the new file next to related code (same directory)
- Distinctive: line 63
MUST fix imports; no dead re-exports “for convenience”
- Distinctive: line 64
MUST keep the diff on this task — no repo-wide re-org
- Distinctive: line 65
NEVER split for SOLID, cleanliness, or “future features”
- Very specific: line 66
NEVER create utils/ / helpers/ / common/ / lib/ for a single function unless that folder is already the repo convention
- Distinctive: line 67
If unsure: don’t split — say so in the summary
- Distinctive: line 73
Before writing code, decide:
- Distinctive: line 79
If the touch list grows past ~5 files for a small task, stop and shrink the plan.
- Distinctive: line 81
If the workspace is empty or has no prior pattern, implement the smallest reasonable solution in-repo.
- Very specific: line 97
If the user explicitly asked for a large scaffold or many files, obey that ask (see User intent wins).
- Distinctive: line 101
Prefer boring, readable code over cleverness
- Distinctive: line 124
When summarizing, say what you changed and what you deliberately did not change.
≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
Kill Slop
Force a minimal, merge-ready diff. Prefer the smallest change that fully solves the stated task.
Hard rules
- NEVER create a file unless the task clearly needs it or a navigability split below applies
- NEVER add a dependency unless the user asked or existing code already requires it
- NEVER refactor unrelated code "while you're here"
- NEVER add comments that restate what the code already says
- NEVER introduce abstractions (wrappers, helpers, factories, "utils") for a single use
- NEVER expand scope beyond the user's request
- NEVER invent TODOs, stubs, or "for later" scaffolding
- NEVER search, list, or read outside the workspace root unless the user explicitly asks
- NEVER browse sibling projects, home, Desktop, or prior chats for “patterns” or inspiration
- NEVER invent Clean Architecture / SOLID folder trees (
domain/,application/,infrastructure/,usecases/, …) unless the repo already uses that layout - MUST match existing project patterns before inventing new ones
- MUST prefer editing an existing file over adding a new one
- MUST delete dead code you introduce; do not leave unused imports/vars
User intent wins
Default to minimal. If the user explicitly asks for more structure, obey that request.
Examples of explicit asks (do what they asked):
- "use Clean Architecture / hexagonal / feature folders"
- "scaffold the full project layout"
- "add dependency X" / "set up Prisma / NextAuth / …"
- "extract a shared abstraction" / "add a utils module"
- "write tests" / "add comments explaining this"
Do not override an explicit ask in the name of kill-slop. Minimalism applies when they did not request the heavier shape. If the ask is ambiguous, prefer minimal and say what you skipped.
File shape (navigability)
Goal: easy to walk for humans and agents — without architecture theater.
Default: colocate. Edit the file/folder that already owns this concern.
Split only when ALL of these hold:
- Pain now — the file is already hard to navigate (rough guide: ≳300–400 lines, or clearly mixed unrelated concerns), and your change would make it worse
- One unit — you’re extracting a single cohesive piece (one component, one route handler, one pure helper) with a clear name
- Local fit — the new file follows an existing nearby pattern (same folder / naming). If there is no pattern, still colocate beside the caller — do not invent a new layer
If only (2) is true (“I could extract this”), do not split.
When you split:
- MUST put the new file next to related code (same directory)
- MUST fix imports; no dead re-exports “for convenience”
- MUST keep the diff on this task — no repo-wide re-org
- NEVER split for SOLID, cleanliness, or “future features”
- NEVER create
utils//helpers//common//lib/for a single function unless that folder is already the repo convention - If unsure: don’t split — say so in the summary
Greenfield: prefer flat under one root (src/ is enough) until real navigability pain appears.
Diff budget
Before writing code, decide:
- Goal — one sentence: what changes for the user
- Touch list — the fewest files that can achieve it
- Non-goals — what you will not do
If the touch list grows past ~5 files for a small task, stop and shrink the plan.
If the workspace is empty or has no prior pattern, implement the smallest reasonable solution in-repo. Do not leave the workspace to find examples.
Measure before finishing
From the project being edited (cwd = that repo), run the bundled read-only checker (never modifies the tree):
bash <path-to-this-skill>/scripts/check-diff.shOptional: --max-files 5 (default), --base HEAD.
STATUS: OK— within soft file budget; still apply the self-check belowSTATUS: OVER_BUDGET— shrink before ending (edit existing files, drop drive-bys)STATUS: SKIP— no git / no base; count files manually with the same ~5-file rule
If the user explicitly asked for a large scaffold or many files, obey that ask (see User intent wins); do not treat OVER_BUDGET as a reason to ignore them.
Style
- Prefer boring, readable code over cleverness
- Handle real error paths the task needs; skip speculative ones
- Names should explain intent; if you need a comment, rename instead
- Tests: only when the repo already tests this area, or the user asks
Self-check (before finishing)
Answer these. If any fails, fix before ending:
- Every changed line serves the stated goal
- No new files that could be inlined into an existing one
- Any new file exists only for navigability (or explicit need) — not architecture cosplay
- New files sit beside related code and match repo layout
- No new deps
- No drive-by renames/moves/refactors
- No obvious comments (
// import x,// return result) - No out-of-workspace reads or “inspiration” fishing
- Diff is something a senior would merge without asking "why is this here?"
- Explicit user asks for architecture / deps / tests / layout were honored (not “killed” away)
- Ran
scripts/check-diff.sh(or manual file count) and addressed OVER_BUDGET if the user did not ask for a large scaffold
Output
When summarizing, say what you changed and what you deliberately did not change. If you split a file, say why (navigability), in one sentence. If the user asked for a heavier shape, say that you followed that ask.
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
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.

