Not in the top 1,000 by GitHub stars
chang416/im-human
讓 AI 用台灣人的繁體中文講話，並把已寫好的文字清掉 AI 味。Claude Code、claude.ai、ChatGPT Codex 與一般聊天都能用。
History
Review weekly changes to SKILL.md before updating. Dates mark the weeks in which we observed each version; checks apply to the recorded commits.
View weekly changes
SKILL.md
Week of Nov 2, 2026 · Line counts unavailable
Before Not checked yet → After Local scripts
First observation → cc845d3d83a2
No earlier version was observed for comparison.
57GitHub stars
1star in 7 days
1skill in this repository
0commits in 30 days
Heat
1.8- Star growth, last 7 daysNot available
- Star growth, last 24 hoursNot available
- Rank movement, last 7 daysNot available
- Commits, last 30 daysNot available
- Accounts asking about it through MCP, last 7 daysNot available
- Hot votes this weekNot available
Heat never changes the order by stars, and sponsorship never touches it. How heat works
Stars over time
30-day change not yet available: not enough history.
Individual skills in this repository
- im-human
/
How it works
im-human
When your agent uses it
台灣繁體中文與英文的語態控制與去 AI 味編輯技能，兩種模式。語態模式：使用者說「我是人」「開啟我是人」「說人話模式」或要 Claude「用我是人講話」時啟動，之後 Claude 自己的每一句輸出都改用台灣人日常的繁體中文（英文則像母語者隨手寫的），禁用 AI 常見的開場白、路標詞、空泛拔高、翻譯腔與簡報式排版，直到使用者說關閉。編輯模式：使用者說「去 AI 味」「說人話」「改自然」「不要像模板」「幫我審稿」，或要改寫貼文、文案、Email、文章、技術文件、小說時啟動，在不更動事實、數字、引用與作者立場的前提下清掉模板感。Also handles English requests to strip AI writing style or make text read like a human wrote it. 只做台灣繁體中文與英文，不支援簡體中文。不用於翻譯、事實查核、憑空創作，也不用於規避 AI 偵測器。
What it loads into context
Frontmatter
285 tokensAlways available to the agent
SKILL.md
3,854 tokensWhen the skill is activated
Reference documents
50,926 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- 語態模式（合約，開啟後每句都適用）
- 編輯模式
- 參考導航（要用才讀，不要全載）
Scripts executed separately
scripts/audit_ai_flavor.py
Files
.gitignoreasset179 bytesLICENSEasset1,065 bytesREADME.mddoc7,852 bytesSKILL.mdcore12,548 bytesSKILL.mdcore12,548 bytesagents/openai.yamlasset370 bytesdocs/promo.mp4asset1,618,298 bytesprompts/chatgpt-custom-instructions.mddoc2,217 bytesprompts/chatgpt-project.mddoc4,808 bytesreferences/bilingual/boundary-cases.mddoc8,984 bytesreferences/bilingual/examples.mddoc12,654 bytesreferences/bilingual/operation-manual.mddoc19,271 bytesreferences/bilingual/phrases-en.mddoc3,861 bytesreferences/bilingual/positive-style.mddoc9,102 bytesreferences/bilingual/protected-spans.mddoc4,800 bytesreferences/bilingual/scene-guardrails.mddoc3,544 bytesreferences/bilingual/scene-packs.mddoc6,043 bytesreferences/bilingual/severity.mddoc4,536 bytesreferences/bilingual/structures.mddoc11,436 bytesreferences/tw/examples.mddoc19,392 bytesreferences/tw/humanize.mddoc6,738 bytesreferences/tw/patterns.mddoc28,393 bytesreferences/tw/protected-list.mddoc3,958 bytesreferences/tw/scenes.mddoc4,691 bytesreferences/tw/taiwan-localization.mddoc5,408 bytesscripts/audit_ai_flavor.pyruns15,256 bytes
Read SKILL.md
# im-human
> 中文觸發詞是「我是人」：使用者說「開啟我是人」「說人話模式」一樣叫得動，Chat 與 Code 都是。
兩件事，先分清楚使用者要哪一件：
| 模式 | 觸發 | 作用對象 |
|---|---|---|
| **語態模式** | 「開啟我是人」「之後都用我是人講話」「說人話模式」，或使用者要你「用這個 skill 講話」 | **你自己接下來的每一句輸出**，直到使用者說「關閉我是人」或「恢復正常」 |
| **編輯模式** | 給你一段文字要你「去 AI 味」「改自然」「審稿」 | 使用者提供的那段文字 |
拿不準就問一句：「要我改這段文字，還是之後都這樣講話？」只問這一次。
---Skills included
- im-human
Repository facts
- License
- MIT
- Created
- Aug 14, 2026
- Last pushed
- Sep 4, 2026
- Forks
- 10
- Open issues
- 0
- Project activity
- Slowing
Before you install
Local scriptsAutomated signals, not a guarantee
Checked Nov 5, 2026
Not fully analyzed: some files were too long, unreadable or missing. This never lowers the level.
Show 1 note
SKILL.md:1Analysis incomplete: Referenced executable target unavailable; not analyzed
- No broad tool permissions
- No dangerous shell commands
- 11 network referencesin 3 files: README.md:52 references/tw/examples.md:33 references/tw/patterns.md:3
- No hidden or obfuscated text
- No instruction overrides
- 1 executable scriptin 1 file: scripts/audit_ai_flavor.py:1
- No access to secrets
- No requests to vote in Hot
Show all 12 signals by file
README.md
- Network accessline 52
git clone https://github.com/chang416/im-human.git "$HOME/.claude/skills/im-human" - Network accessline 58
git clone https://github.com/chang416/im-human.git "$env:USERPROFILE\.claude\skills\im-human" - Network accessline 89
到 [claude.ai 的 Customize → Skills](https://claude.ai/customize/skills) 上傳， - Network accessline 155
`references/tw/patterns.md` 那份 38 種痕跡清單，整理自[中文維基百科「AI生成文的特徵」](https://zh.wikipedia.org/zh-tw/Wikipedia:AI生成文的特徵)、朱宥勳的「AI腔」句型分析，以及英文維基的 Signs of AI writing，範例全部針對台灣的內容創作與辦公場景重寫。每一種痕跡都附「誤殺邊界」，因為問題從來不是句型本身，是密度，以及底下有沒有東西撐著。
references/tw/examples.md
- Network accessline 33
> 與其在原地焦慮，不如現在就出發。畢竟，機會永遠留給準備好的人，不是嗎？🔥 > > 👉 立即報名：https://example.com/signup?utm_source=chatgpt.com - Network accessline 35
> 👉 立即報名：https://example.com/signup?utm_source=chatgpt.com - Network accessline 45
> 我自己用這套流程寫電子報，每期少花一個半小時。省下來的時間拿去幹嘛？睡覺，真的。 > > 報名：https://example.com/signup - Network accessline 47
> 報名：https://example.com/signup
references/tw/patterns.md
- Network accessline 3
> 主要整理自[中文維基百科「AI生成文的特徵」](https://zh.wikipedia.org/zh-tw/Wikipedia:AI生成文的特徵)（WikiProject AI Cleanup 社群持續更新）、朱宥勳「AI腔」句型分析，以及英文維基「Signs of AI writing」的社群觀察，範例句全部針對繁體中文的內容創作與辦公場景重寫。 - Network accessline 25
**二、形式正確不等於內容正確。** 在 AI 之前，能把文章形式寫對的人，內容水準通常也在；AI 打破了這個對應關係，現在亂七八糟的內容也能被裝得人模人樣（[朱宥勳的觀察](https://www.youtube.com/watch?v=9uuX6cb81C8)）。對改寫的人來說這有兩層意思：讀稿時不要因為「讀起來很順、很有邏輯感」就放過它；改稿時也不要以為把 38 種痕跡清乾淨了，這篇文章就有想法了。**清痕跡處理的是形式，想法要作者自己有。** - Network accessline 142
**先講清楚：這個句型本身沒有錯。** 它做的是「定義」和「區分」兩個寫作的基本動作，英美學術寫作訓練本來就教這個（[朱宥勳的觀察](https://www.youtube.com/watch?v=9uuX6cb81C8)）。問題出在它被用到浮濫，濫用到只要一出現就被讀者判定是 AI 寫的——一個好用的句型被玩壞了。
scripts/audit_ai_flavor.py
- Executable scriptline 1
#!/usr/bin/env python3
How the signals fit together
Facts from a static read of the files at the pinned commit. They describe what the code does, not why.
Signal confidence: 12 facts
- Network access in code of documentationREADME.md:52Low confidence
- Network access in code of documentationREADME.md:58Low confidence
- Network access in prose of documentationREADME.md:89Low confidence
- Network access in prose of documentationREADME.md:155Low confidence
- Network access in prose of documentationreferences/tw/examples.md:33Low confidence
- Network access in prose of documentationreferences/tw/examples.md:35Low confidence
- Network access in prose of documentationreferences/tw/examples.md:45Low confidence
- Network access in prose of documentationreferences/tw/examples.md:47Low confidence
- Network access in prose of documentationreferences/tw/patterns.md:3Low confidence
- Network access in prose of documentationreferences/tw/patterns.md:25Low confidence
- Network access in prose of documentationreferences/tw/patterns.md:142Low confidence
- Executable script in code of a file that runsscripts/audit_ai_flavor.py:1High confidence
Works with
| Agent | Status | Evidence | Install to |
|---|---|---|---|
| Claude Code | Compatible | Standard SKILL.md SKILL.md | .claude/skills/<skill-name>/ |
| Codex | Compatible | Standard SKILL.md SKILL.md | .agents/skills/<skill-name>/ |
| Cursor | Compatible | Standard SKILL.md SKILL.md | .cursor/skills/<skill-name>/ |
| Gemini CLI | Compatible | Standard SKILL.md SKILL.md | .gemini/skills/<skill-name>/ |
| GitHub Copilot | Compatible | Standard SKILL.md SKILL.md | .github/skills/<skill-name>/ |
| OpenCode | Compatible | Standard SKILL.md SKILL.md | .opencode/skills/<skill-name>/ |
- Native The repository ships a manifest, folder or frontmatter field for this agent.
- Compatible A standard SKILL.md that this agent can read.
- Unknown No evidence either way.
Derived automatically from paths, manifests and frontmatter. Not tested on each agent.
