Meiguang Car Insurance

车险保单PDF字段提取工具 v5.0.2（美光/大地/太平洋/浙商/亚太/人保/华海）。输入PDF文件夹，输出Excel。支持14字段，含车辆使用性质多数纠正、PDAA/PDZA表格兜底、Excel中文UTF-8编码修复。

This package appears to be what it says: a PDF-to-Excel extractor specialized for certain Chinese insurers. Before running it: 1) Inspect scripts/run_extract.py (and any helper it imports) to confirm which files and directories it reads/writes and to ensure there are no network calls or unexpected behaviors. 2) Because several debug scripts contain hard-coded absolute Windows paths (C:\Users\Administrator\Desktop and .openclaw workspace), run it in a sandbox or on a non-production machine and correct paths as needed. 3) Ensure required Python dependencies (pymupdf, pdfplumber, openpyxl, etc.) are installed in an isolated virtualenv/container; the skill does not declare or install them. 4) Be aware the tool extracts sensitive personal data from PDFs (names, ID numbers, phone numbers, VINs); protect any processed files and outputs. 5) If you need higher assurance, post the full run_extract.py for review (it’s large and was not fully shown here) to confirm there are no network transmissions, credential usage, or other unexpected actions. Running with a least-privileged user or in a VM/container is recommended.

Type: OpenClaw Skill Name: meiguang-car-insurance Version: 5.1.1 The skill bundle is a specialized tool for extracting structured data from various Chinese car insurance PDF policies into Excel format. It utilizes standard libraries such as pdfplumber, pymupdf (fitz), and pandas. While the bundle includes numerous diagnostic and 'fix' scripts (e.g., run_extract.py, fix_chinese_num.py) and relies on hardcoded absolute Windows file paths (targeting the 'Administrator' desktop), these behaviors are consistent with a developer's local workspace for document parsing. There is no evidence of malicious intent, data exfiltration, or unauthorized execution of remote code.