humanizer-de 安全扫描报告 — 可信 | ClawSafe

5 /100

humanizer-de

Deutscher KI-Text-Detektor – 5-Pass-Analyse, 24 Muster, 168 Vokabeln, 5 statistische Signale, Personality Injection

Humanizer-DE is a legitimate German AI-text detector and humanizer with no malicious behavior — it performs offline text analysis using only fs/path, declares no network access, and contains no credential theft, exfiltration, or obfuscation.

技能名称humanizer-de

分析耗时28.0s

引擎pi

✓

可以安装

Approve for deployment. The skill operates entirely offline and poses no security risk.

资源类型	声明权限	推断权限	状态	证据
文件系统	`READ`	`READ`	✓ 一致	SKILL.md declares offline-only; JS script uses fs.readFileSync/writeFileSync for…
网络访问	`NONE`	`NONE`	—	SKILL.md rule 5: 'Kein Netzwerk'; JS script has zero require('http')/fetch/axios…
命令执行	`NONE`	`NONE`	—	No subprocess, exec, or child_process in humanize-de.js
环境变量	`NONE`	`NONE`	—	No process.env access in any script
技能调用	`READ`	`READ`	✓ 一致	SKILL.md describes 6 LLM-based commands; _meta.json declares skill_invoke
剪贴板	`NONE`	`NONE`	—	No clipboard access found
浏览器	`NONE`	`NONE`	—	No browser/DOM APIs used
数据库	`NONE`	`NONE`	—	No database access

3 项发现

🔗

中危外部 URL 外部 URL

https://tikitackr.github.io/OpenClaw-Dashboard/

README.md:140

🔗

中危外部 URL 外部 URL

https://tikitackr.github.io/Cowan/

README.md:141

🔗

中危外部 URL 外部 URL

https://clawhub.ai/Tikitackr/humanizer-de

_meta.json:28

目录结构

12 文件 · 133.7 KB · 3447 行

Markdown 10f · 2458L JavaScript 1f · 960L JSON 1f · 29L

├─ ▾ 📁 _dev

│ └─ 📝 SESSION-LOG.md Markdown 215L · 8.6 KB

├─ ▾ 📁 references

│ ├─ ▾ 📁 stil-layer

│ │ ├─ 📝 basis.md Markdown 110L · 4.9 KB

│ │ └─ 📝 lesch.md Markdown 175L · 7.4 KB

│ ├─ 📝 examples.md Markdown 279L · 14.2 KB

│ ├─ 📝 ki-muster.md Markdown 446L · 16.1 KB

│ ├─ 📝 personality-injection.md Markdown 259L · 10.0 KB

│ ├─ 📝 statistische-signale.md Markdown 316L · 10.1 KB

│ └─ 📝 vokabeln.md Markdown 278L · 16.3 KB

├─ ▾ 📁 scripts

│ └─ 📜 humanize-de.js JavaScript 960L · 29.4 KB

├─ 📋 _meta.json JSON 29L · 893 B

├─ 📝 README.md Markdown 147L · 6.0 KB

└─ 📝 SKILL.md Markdown 233L · 9.7 KB

依赖分析 1 项

包名	版本	来源	已知漏洞	备注
`node (fs, path builtins only)`	`>=18`	stdlib	否	Zero npm dependencies; only fs and path from Node.js stdlib

安全亮点

✓ No network access — skill explicitly declares 'Kein Netzwerk' and JS script contains zero HTTP/client libraries

✓ No shell execution — humanize-de.js uses only fs and path from Node.js stdlib

✓ No credential access — no process.env, ~/.ssh, .env, or sensitive path access anywhere

✓ No obfuscation — all code is plain JavaScript with readable variable names and comments

✓ No data exfiltration — results are printed to stdout or written to local files only

✓ Open source — all code is readable, no compiled or packed payloads

✓ No external dependencies — JS script uses only Node.js built-ins (fs, path)

✓ Clear documentation — SKILL.md accurately describes all 6 commands and 5-pass workflow

✓ Session-Log confirms no code was copied from external operator-humanizer ZIP

✓ _meta.json explicitly declares requires.network: false