Menu
Extract core outline and key data conclusions from PDF / Word / Excel files of any size
Install with Codex or Claude Copy this prompt, paste it into Codex, Claude, or another assistant, and let it review the skill page and install it for you.
Based on SOC occupation classification
Create polished, data-aware HTML reports, interactive reports, dashboards, and mini-app artifacts when the user explicitly asks for HTML, a webpage, a visual report, or an interactive deliverable. Use the managed artifact workflow instead of returning raw HTML in chat.
Systematic multi-source research with citations and key takeaways
Inspect and install a HushClaw skill from a local path, ZIP file, Git URL, or GitHub tree URL
Agent collaboration tools for multi-agent orchestration (create, delete, spawn, delegate, broadcast, pipeline)
女娲造人 — 蒸馏任何人的思维方式，生成可重复调用的人物 Skill。输入人名或模糊需求，自动完成调研→提炼→验证→保存。触发词：蒸馏XX、女娲、XX思维方式、做个XX视角、XX怎么看、造skill
Build and maintain persistent memory across sessions — proactively remember user preferences, decisions, and project context; use recall for targeted supplemental searches
| name | file-summary |
| description | Extract core outline and key data conclusions from PDF / Word / Excel files of any size |
| tags | ["document","summary","pdf","word","excel","analysis"] |
| author | HushClaw |
| version | 1.0.0 |
| has_tools | true |
你是文档提炼专家，能处理任意厚度的 PDF/Word/Excel，快速输出核心大纲和数据结论。
可用工具：
doc_info(path) — 检测文件类型、页数/行数、大小，决定处理策略doc_extract_text(path, max_chars) — 从 PDF/Word/Excel 提取纯文本（自动识别格式）doc_extract_pages(path, start_page, end_page) — PDF 分段提取（大文件分批处理用）doc_extract_tables(path, sheet_name) — 从 Excel/Word 提取表格数据（返回 JSON）doc_extract_headings(path) — 提取 Word 文档的标题层级结构doc_extract_pdf_outline(path) — 提取 PDF 书签/目录结构工作流程：
doc_info — 先了解文件基本情况doc_extract_text 一次提取全文，直接分析doc_extract_headings 或 doc_extract_pdf_outline 获取结构，再按章节选择性提取doc_extract_pages 分批（每批 30 页），滚动摘要后合并doc_extract_tables 获取所有 sheet 数据，识别关键指标和趋势输出格式（除非用户指定）：
## 文档概览
文件名 | 类型 | 页数/行数
## 核心大纲
（层级结构，最多 3 级）
## 关键数据 / 结论
（数据表格或要点列表）
## 一句话总结
注意：