anti-slop

Skills that stop AI coding agents from shipping garbage.

One command:

npx skills add iCodeCraft/anti-slop

Same prompt → smaller diffs. Less fake architecture. PRs you’d actually merge.

Works with Cursor, Claude Code, Codex, and other agents via Agent Skills.

---

Same prompt. Half the slop.

Empty folder. Same model. This prompt:

Create a production-ready FastAPI backend for authentication and a personal todo list.
I need register, login, create/list/toggle todos. In-memory storage is fine.

	Without	With kill-slop
Files	13	5
Lines	433	222

JWT auth and todos either way. anti-slop cuts the architecture cosplay — routers, schemas, dependencies, config theater → main + auth + store.

Reproduce the A/B: examples/README.md.

---

The problem

Agents overbuild. Extra files. Strategy patterns. Obvious comments. Drive-by refactors.

You correct the same junk every chat. anti-slop writes the rules down once.

---

Install

npx skills add iCodeCraft/anti-slop

npx skills add iCodeCraft/anti-slop -a cursor -y
npx skills add iCodeCraft/anti-slop -a claude-code -y
npx skills add iCodeCraft/anti-slop -g -y

New agent session → skills load when relevant, or call /kill-slop, /security-review, /pr-hygiene.

npx skills list
npx skills remove kill-slop

---

Skills

Skill	Does
kill-slop	Minimal diffs. No drive-by refactors, junk comments, or unrequested files/deps
security-review	Diff review: authz, injection, secrets, SSRF, unsafe defaults
pr-hygiene	Tight scope, honest PR body, no leftover debug

---

How it works

Each skill is a folder with a SKILL.md (Agent Skills). Agents load name + description cheaply, then the full playbook when it matches the task.

Tool	Project	Global
Cursor	.agents/skills/	~/.cursor/skills/
Claude Code	.claude/skills/	~/.claude/skills/
Codex / others	.agents/skills/	under ~/

Install via the skills CLI.

License

MIT