Agent Skills · Office Efficiency
Curated skills for Claude, Cursor and other AI agents. Copy one prompt to install.
Uses the native macOS Vision framework to extract Simplified Chinese, Traditional Chinese, and English text from images, screenshots, or scans, outputting each text block to stdout.
Upload two contract versions to align paragraphs, highlight additions, deletions, and changes, and generate a severity-sorted structured diff report without legal advice.
Extract feature hierarchy from Word, PDF, text, and Markdown documents into a structured Excel feature list with descriptions.
Recognize text, tables, and layout from images, PDFs, and scanned documents via Xiangyun OCR, with exports to Excel, Word, and Markdown.
Create, modify, and analyze DOCX documents by generating them with JavaScript or by unpacking to edit the raw XML.
Plans PPT content, structure, style, and copy, then coordinates create-ppt to produce the final deck.
Use an App Key to interpret tenders, generate bids, and audit compliance risks.
Judges whether meetings should be held, trimmed, moved async, or canceled, with scoring, time-savings estimates, and fallback paths.
Provides practical Word, Excel, PPT, WPS, and automation guidance, including layout, formulas, design, and office scripts.
Reviews Chinese official documents across format, citation sourcing, wording, numbers, style, logic, and ambiguity, producing a severity-graded proofreading report.
Polishes email wording, structure, and tone in Chinese or English, offers formal, friendly, or urgent versions, and checks grammar without sending emails.
Convert a single local Office, image, text, or HTML file to PDF via Duhui's async API, with temporary OSS upload, polling, and download, requiring DUHUI_ALI_APPCODE.