shuorenhua
shuorenhua ('Speak Human') by MrGeDiao is an open-source, highly targeted prompt engineering skill (MIT license) built to eliminate the notorious 'AI accent'—the bloated, repetitive, and synthetic phrasing that plagues Chinese large language model completions.
Installable across Claude Code, Codex, Cursor, and standard LLM chat interfaces via `npx skills add MrGeDiao/shuorenhua`, the skill acts as an uncompromising copy editor. It systematically removes empty greetings, repetitive preamble clauses, and hollow transition markers (such as '值得注意的是', '总的来说', '在当今快节奏的世界中'), condensing verbose drafts into punchy, professional prose while strictly guaranteeing that real numbers, logical conditions, commitments, and the author's conversational voice remain untouched.
For developers and knowledge workers producing weekly status reports, technical release notes, or internal memos, shuorenhua turns synthetic drafts into clean, ready-to-publish communication. However, as an editing heuristic, it is tailored primarily for practical, factual business and engineering communication, and is unsuited for creative literature or expressive poetry where repetitive rhythm may be intentional.
Features
Why choose it
Trade-offs
Maintenance check
- Last commit
- 2026-11-10
- Commits, last 90 days
- 32
- Releases, last 90 days
- 8
- Latest release
- v2.7.0 (2026-11-10)
- Open issues
- 1
- Recent issues with at least one comment
- 12/12
Oldest week on the left; the latest four weeks are orange. Source: GitHub commit activity.
Status rule: archived on GitHub = Archived; last commit within 90 days = Active; 91 to 365 days = Slowing; older = Stale. All figures come from the GitHub API on the date shown.
Most discussed open issues
- #5 征集 bad case:改完还是像 AI?贴出来,我拿真实样本改规则5 comments · 2026-06-10
Issue titles are quoted from GitHub. Open the link to check whether a problem still applies to the version you use.
Compatibility
Quick start
Use cases
Alternatives
Comments
No comments yet. Be the first!