MrGeDiao/shuorenhua
说人话 — 让 AI 写出来的东西像人写的。 | Bilingual AI writing de-slop skill for Codex, Openclaw, Claude Code, Cursor, ChatGPT
A rule-based system with 210+ Chinese phrases, 96 English phrases, and 19 structural anti-patterns organized across six categories (corporate jargon, engineer-speak, Xiaohongshu style, translation artifacts, etc.), deployed via prompt injection into Claude Code, Cursor, Codex, and ChatGPT. It applies three severity tiers and context-aware filtering (four scenario modes: chat, technical summaries, docs, blogs) with safeguards against false positives in technical terms, citations, and system prompts. The evaluation benchmark shows 100% accuracy on 37 test cases with zero false positives, and includes an annotation-only mode for review workflows.
Stars
42
Forks
4
Language
—
License
MIT
Category
Last pushed
Mar 26, 2026
Commits (30d)
0
Get this data via API
curl "https://pt-edge.onrender.com/api/v1/quality/prompt-engineering/MrGeDiao/shuorenhua"
Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.
Higher-rated alternatives
maxvaega/skillkit
Implementing Skills functionality for your agents
nidhinjs/prompt-master
A Claude skill that writes the accurate prompts for any AI tool. Zero tokens or credits wasted....
telagod/code-abyss
☠️ 一键为 Claude Code / Codex CLI 注入邪修人格与 40+ 安全工程秘典 | npx code-abyss
Imbad0202/academic-research-skills
Academic Research Skills for Claude Code: research → write → review → revise → finalize
ursisterbtw/ccprompts
practical claude code commands and subagents