It is designed for scenarios that require direct operating system application and in-depth data analysis. [Forced trigger scenario]: - User mentions reading/writing/manipulating Excel, WPS, Word, TXT, Markdown, RTZ, etc. - User wants to "grab", "extract", and "get" data from any application - User needs to perform "in-depth analysis", "trend research", "anomaly detection", and "prediction" on the data - User requests to generate "charts", "visualizations", "dashboards", "data reports" - users say, "Help me see in this document..." Analyze this data...", "Make a chart presentation..." - Any task involving cross-application data flow [Core Competencies]: System interface calls × Data in-depth analysis × Professional visualization IMPORTANT: As long as it involves any of the file operations, data analysis, and visualization, this skill must be used. Don't skip tasks just because they "look simple" - there are many pitfalls in the underlying interface calls, and there are pitfall avoidance guides in the skills.

专为需要直接操作系统应用并进行深度数据分析的场景设计。 【强制触发场景】： - 用户提及 Excel、WPS、Word、TXT、Markdown、RTZ 等文件的读取/写入/操控 - 用户想从任何应用中「抓取」「提取」「获取」数据 - 用户需要对数据进行「深度分析」「趋势研究」「异常检测」「预测」 - 用户要求生...

Security Scan

The skill's code and instructions match its stated purpose (local file reading, deep analysis, visualization), but there are important inconsistencies and autonomy risks you should review before installing or enabling it.

Purpose & Capability

The skill's name/description (system-level file/app access + deep analysis + visualization) aligns with the provided scripts (win/mac/linux readers, analyzer, viz engine). However the registry metadata declares no required binaries/configs/env yet the SKILL.md and scripts rely on system components (Excel/WPS via COM, xlwings/AppleScript, libreoffice, osascript) and external Python packages. The omission of required system dependency declarations (Office/LibreOffice, macOS Accessibility/osascript) is an incoherence that could lead to unexpected failures or overbroad agent behavior.

Instruction Scope

SKILL.md and AGENTS.md instruct the agent to auto-detect OS and automatically read, analyze, visualize, and save reports for 'any' supported file; they explicitly say this skill must be used for all file/data analysis tasks and discourage asking the user about format. That encourages agents to open arbitrary local files and perform full analysis without explicit per-file confirmation. The runtime instructions also recommend granting macOS Accessibility/automation permissions and running system commands (osascript, libreoffice), which are legitimate for the use case but broaden the scope of actions the agent will perform.

Install Mechanism

No formal install spec in registry (instruction-only), but the repository includes code and a README with a git clone URL and requirements.txt. Dependencies are standard PyPI packages; no downloads from unusual hosts in the provided files. The lack of an explicit install/packaging declaration in the skill metadata is a minor mismatch but not necessarily malicious—just means installation and dependency management are left to the operator.

Credentials

The skill requires no environment variables or external credentials (correctly reflected in metadata). However it legitimately requires access to local system resources (Office COM objects, AppleScript automation, LibreOffice binary) and may ask user to grant macOS Accessibility or allow terminal to control Excel. Those are proportional to the stated purpose but are privacy-sensitive because the skill will read arbitrary local documents and write outputs and logs to disk.

Persistence & Privilege

The skill does not request always:true, does not declare special platform-wide privileges, and does not modify other skills’ configs. README suggests placing the repo under ~/.cursor/skills for automatic loading (normal for skills). No evidence the skill persistently exfiltrates credentials or alters system-wide settings beyond saving outputs/logs in an outputs/ directory.

What to consider before installing

This skill appears to do what it says (read local Office/WPS files, analyze data, create visual reports), but it will operate on local files and call system automation (COM, AppleScript, LibreOffice). Before installing: 1) Review and trust the code and its GitHub source — it will read arbitrary files you point it at. 2) Be aware it asks for/needs system-level permissions on macOS (Accessibility/automation) and requires Office/LibreOffice to be present for some paths. 3) The skill metadata omits declaring these system dependencies; verify and install required packages (requirements.txt) and binaries manually. 4) If you care about privacy, run it in a controlled environment (VM or sandbox) or require explicit user confirmation before the agent opens files. 5) If you will allow autonomous agent invocation, consider limiting triggers or requiring an explicit user approval step, because SKILL.md's “must use” language encourages automatic analysis of files without additional prompts.

System Data Intelligence Skill

🔍 快速决策树

收到任务时，先判断类型：

用户任务
  ├─ 涉及文件/应用操作？
  │    ├─ Windows → 走 [WIN-PATH]
  │    ├─ macOS   → 走 [MAC-PATH]
  │    └─ Linux   → 走 [LINUX-PATH]
  ├─ 纯数据分析（已有数据）？→ 走 [ANALYSIS-PATH]
  └─ 数据可视化？→ 走 [VIZ-PATH]

遇到组合型任务（最常见）：先读文件 → 再分析 → 最后可视化。

---

[WIN-PATH] Windows 系统接口调用

优先级顺序

1. Python + win32com（Excel/Word/WPS 首选，功能最全）
2. PowerShell + COM（系统级操作，无需 Python 环境）
3. openpyxl / python-docx（离线解析，不依赖应用程序）
4. pywinauto（GUI 自动化，万不得已时使用）

Excel / WPS Spreadsheet

# 执行: python scripts/win_excel_reader.py <filepath> [sheet_name]

Word / WPS Writer

# 执行: python scripts/doc_parser.py <filepath>

关键注意点

• 操作完成后必须调用 Quit() 释放 COM 进程
• WPS 的 COM 名称："KET.Application"（表格）/ "KWPS.Application"（文字）
• 大文件（>50MB）使用 openpyxl 的 read_only=True 模式

详细 API 手册 → references/windows-api.md

---

[MAC-PATH] macOS 系统接口调用

优先级顺序

1. Python + xlwings（Excel for Mac 首选）
2. AppleScript / JXA（系统原生，稳定可靠）
3. python-docx / openpyxl（离线解析）
4. subprocess + osascript（调用系统命令）

Excel for Mac

# 执行: python scripts/mac_excel_reader.py <filepath> [sheet_name]

关键注意点

• 首次调用 AppleScript 需要「辅助功能」授权
• 提示用户在「系统设置 → 隐私与安全性 → 辅助功能」开启权限

详细 API 手册 → references/macos-api.md

---

[LINUX-PATH] Linux 系统接口调用

优先级顺序

1. openpyxl / python-docx（Excel/Word 首选，无需应用程序）
2. LibreOffice headless（老格式 .doc/.xls 转换）
3. pandas + xlrd（旧版 Excel 离线解析）
4. pdfplumber / pymupdf（PDF 提取）

Excel / CSV 数据

# 执行: python scripts/doc_parser.py <filepath>
# 支持 .xlsx .xls .xlsm .csv （无需 Office）

老格式转换（.doc / .xls）

# LibreOffice headless 转换为现代格式
libreoffice --headless --convert-to xlsx input.xls --outdir /tmp/
libreoffice --headless --convert-to docx input.doc --outdir /tmp/

关键注意点

• Linux 无 COM/AppleScript，一律用 Python 库离线解析
• 老格式文件先用 LibreOffice headless 转换再读取
• 中文内容需要安装 CJK 字体：sudo apt install fonts-noto-cjk

详细 API 手册 → references/linux-api.md

---

[ANALYSIS-PATH] 数据深度分析

分析层次（从浅到深）

Level 1: 描述性分析   → 数据现状是什么？（均值、分布、缺失率）
Level 2: 诊断性分析   → 为什么会这样？（相关性、异常根因）
Level 3: 预测性分析   → 未来会怎样？（趋势、预测模型）
Level 4: 规范性分析   → 应该怎么做？（优化建议、决策支持）

标准分析流水线

# 执行完整分析
# python scripts/deep_analyzer.py <csv_or_excel_path> [date_col] [value_col]

脚本输出：

• outputs/analysis_result.json — 结构化分析报告
• outputs/summary.md — 文字洞察摘要

详细分析模式 → references/viz-patterns.md

---

[VIZ-PATH] 数据可视化

图表选型

数据关系类型
  ├─ 时间趋势        → 折线图 / 面积图
  ├─ 类别比较        → 柱状图 / 条形图 / 雷达图
  ├─ 部分与整体      → 饼图 / 旭日图 / 树状图
  ├─ 分布情况        → 箱线图 / 直方图 / 小提琴图
  ├─ 相关关系        → 散点图 / 热力图
  └─ 多维关系        → 平行坐标 / 桑基图

可视化执行

# 生成交互式仪表盘
# python scripts/viz_engine.py <analysis_result.json> <output_dir>

输出：report.html（交互版）+ charts/*.png（静态版）

图表模板与最佳实践 → references/viz-patterns.md

---

⚡ 自动文件格式检测

不确定文件格式时，使用统一入口：

from scripts.doc_parser import detect_and_load
df = detect_and_load("/path/to/any/file")

支持格式：.xlsx .xls .xlsm .et .docx .doc .wps .txt .md .rtz .csv .json

---

📤 输出规范

每次任务完成必须输出：

1. 数据摘要卡片（≤5 行关键洞察，Markdown 格式）
2. 可视化文件（HTML 交互版 + PNG 静态版）
3. 结构化数据（JSON / CSV，供后续使用）
4. 操作日志（记录调用接口与数据量）

输出路径：outputs/report_YYYYMMDD_HHMMSS/

---

💡 心法

不要问用户想要什么格式——直接给最好的那个。 收到文件就分析，分析完就可视化，可视化完就生成报告。 每一步都留下日志，每一步都输出可下载文件。 用户说「分析一下」，你就给他一份完整的数据故事。