markdown-multilingual-translator 安全扫描报告 — 可信 | ClawSafe

5 /100

markdown-multilingual-translator

A specialized tool for translating Markdown files while preserving structural elements, code blocks, links, and formatting. Supports six languages: English, Simplified Chinese, Taiwan Traditional Chinese, Japanese, Korean, and Indonesian.

This is a legitimate Markdown translation tool with no malicious behavior detected. All operations are appropriate for the stated functionality.

技能名称markdown-multilingual-translator

分析耗时26.2s

引擎pi

✓

可以安装

This skill is safe to use. No security concerns identified.

资源类型	声明权限	推断权限	状态	证据
文件系统	`READ`	`WRITE`	✓ 一致	translate_markdown.py:122-127 - reads input file, writes output file as declared…
网络访问	`READ`	`READ`	✓ 一致	translator.py:73 - makes POST requests to api.anthropic.com for Claude API trans…

2 项发现

🔗

中危外部 URL 外部 URL

https://api.example.com/users

references/EXAMPLES.md:237

🔗

中危外部 URL 外部 URL

https://api.example.com

references/EXAMPLES.md:381

目录结构

13 文件 · 152.1 KB · 4886 行

Markdown 6f · 2593L Python 6f · 1884L JSON 1f · 409L

├─ ▾ 📁 references

│ ├─ ▾ 📁 glossaries

│ │ └─ 📋 tech_glossary_base.json JSON 409L · 12.0 KB

│ ├─ 📝 EXAMPLES.md Markdown 646L · 14.0 KB

│ ├─ 📝 GLOSSARY_TEMPLATE.md Markdown 486L · 13.7 KB

│ └─ 📝 LANGUAGE_GUIDE.md Markdown 535L · 18.6 KB

├─ ▾ 📁 scripts

│ ├─ 🐍 language_detector.py Python 204L · 6.4 KB

│ ├─ 🐍 markdown_parser.py Python 421L · 13.5 KB

│ ├─ 🐍 terminology_manager.py Python 336L · 10.5 KB

│ ├─ 🐍 translate_markdown.py Python 302L · 10.2 KB

│ ├─ 🐍 translator.py Python 334L · 10.8 KB

│ └─ 🐍 validator.py Python 287L · 10.4 KB

├─ 📝 ARCHITECTURE.md Markdown 174L · 6.5 KB

├─ 📝 README.md Markdown 354L · 12.1 KB

└─ 📝 SKILL.md Markdown 398L · 13.4 KB

依赖分析 3 项

包名	版本	来源	已知漏洞	备注
`markdown`	`not specified`	pip	否	Standard Markdown parsing library
`pyyaml`	`not specified`	pip	否	Standard YAML parsing for frontmatter
`requests`	`not specified`	pip	否	Standard HTTP library for API calls

安全亮点

✓ No credential theft - API key is used only for intended Claude API calls

✓ No data exfiltration - all network calls are to legitimate Anthropic API endpoint

✓ No obfuscation - all code is readable plain Python

✓ No sensitive path access - only operates on user-specified input/output files

✓ Standard dependencies - markdown, pyyaml, requests are well-known, reputable libraries

✓ No hidden functionality - all operations match stated documentation

✓ No reverse shell or C2 communication detected

✓ No base64-encoded payloads or eval() calls

✓ Batch processing is limited to .md files as documented

✓ Glossary loading is file-path based with proper error handling