可信 — 风险评分 0/100
上次扫描:2 天前 重新扫描
0 /100
expense-reimbursement
Travel Expense Reimbursement Assistant - Automates receipt scanning, classification, form filling, and print package generation for Chinese corporate expense reports
This is a legitimate travel expense reimbursement assistant with comprehensive documentation, user confirmation checkpoints, and well-defined file/document processing capabilities. No malicious behavior detected.
技能名称expense-reimbursement
分析耗时29.1s
引擎pi
可以安装
This skill is safe to use. No action required.
资源类型声明权限推断权限状态证据
文件系统 WRITE WRITE ✓ 一致 SKILL.md lines 1-50: Full workflow documents file read/write for PDFs, DOCX, ima…
网络访问 READ READ ✓ 一致 SKILL.md: XSD schema references for XML parsing; scripts/unzip_and_parse.py line…
命令执行 READ READ ✓ 一致 SKILL.md: Documents tesseract (Linux/Windows) and qlmanage (macOS) for OCR
环境变量 READ READ ✓ 一致 scripts/unzip_and_parse.py: REIMBURSEMENT_DIR env var documented and declared in…
技能调用 NONE NONE No skill invocation detected
剪贴板 NONE NONE Not used
浏览器 NONE NONE Not used
数据库 NONE NONE Not used
2 项发现
🔗
中危 外部 URL 外部 URL
http://invoice.xsd
scripts/unzip_and_parse.py:84
🔗
中危 外部 URL 外部 URL
http://www.typing/xsd
scripts/unzip_and_parse.py:84

目录结构

3 文件 · 35.9 KB · 1004 行
Markdown 2f · 799L Python 1f · 205L
├─ 📁 references
│ └─ 📝 workflow.md Markdown 95L · 2.7 KB
├─ 📁 scripts
│ └─ 🐍 unzip_and_parse.py Python 205L · 8.1 KB
└─ 📝 SKILL.md Markdown 704L · 25.1 KB

依赖分析 4 项

包名版本来源已知漏洞备注
python-docx latest pip Used for Word form reading and filling - legitimate document processing
pypdf latest pip PDF merging - legitimate document processing
reportlab latest pip Image-to-PDF conversion with A4 scaling - legitimate document processing
Pillow latest pip Image dimension reading - legitimate document processing

安全亮点

✓ Comprehensive SKILL.md with 704 lines of clear documentation
✓ User confirmation checkpoints at Steps 5.5, 7.1, 7.3, 7.4, and 8 - prevents autonomous operation
✓ Original file protection: 00_原始资料/ always retained, deletions require explicit authorization
✓ Template requirement enforced: skill pauses if template missing, cannot skip
✓ Project code confirmation required before form filling
✓ All dependencies documented: python-docx, pypdf, reportlab, Pillow, tesseract/qlmanage
✓ Clean Python implementation using only standard libraries (zipfile, os, xml.etree.ElementTree)
✓ No credential harvesting or sensitive path access
✓ No external network requests or data exfiltration
✓ Legitimate document processing: PDF merge, image-to-PDF conversion, DOCX form filling
✓ OCR tools (tesseract/qlmanage) properly documented as dependencies
✓ Supports cross-platform (macOS/Linux/Windows) with proper platform detection
✓ Configurable paths through environment variables (documented)
✓ Automatic skipping of irrelevant files (.DS_Store, 申请截图.png)
✓ Recursive ZIP scanning properly declared for receipt processing