document-pro

Document Pro - Document processing skill

Overview

Empower AI with powerful document processing capabilities:

• PDF reading and extraction
• Word document parsing
• PowerPoint extraction
• Excel data extraction
• Document format conversion

Trigger scenarios

1. The user sends a document and asks to "analyze" or "summarize" it
2. The user asks to "extract document content"
3. The user asks to "convert to PDF"
4. The user asks about specific information in a document
5. The user asks to "extract key points from a report/paper"

Supported formats

Format	Read	Write	Tool
PDF	✅	✅	pdfplumber, PyPDF2
DOCX	✅	✅	python-docx
PPTX	✅	❌	python-pptx
XLSX	✅	✅	openpyxl
TXT	✅	✅	Built-in
Markdown	✅	✅	Built-in

Tool usage

PDF processing

# 提取文本
import pdfplumber

with pdfplumber.open("document.pdf") as pdf:
    for page in pdf.pages:
        text = page.extract_text()
        print(text)

# 提取表格
with pdfplumber.open("document.pdf") as pdf:
    table = pdf.pages[0].extract_tables()

Word documents

from docx import Document

doc = Document("document.docx")
for para in doc.paragraphs:
    print(para.text)

# 提取表格
for table in doc.tables:
    for row in table.rows:
        print([cell.text for cell in row.cells])

PowerPoint

from pptx import Presentation

prs = Presentation("presentation.pptx")
for slide in prs.slides:
    for shape in slide.shapes:
        if shape.has_text_frame:
            print(shape.text)

Workflow

1. 识别文档类型 → 选择正确的工具
2. 读取内容 → 提取文本、表格、图片
3. 分析信息 → 理解结构、提取要点
4. 总结呈现 → 用中文总结给用户

Advanced features

Document summarization

• Extract the main points of the document
• Generate a brief summary
• List key points

Table processing

• Identify table structure
• Extract table data
• Convert to CSV/Excel

Keyword extraction

• Find important nouns/terms
• Identify topics
• Extract key information

Output format

When presenting a document to the user:

• Document type and page count
• Main content summary
• Key points (3-5 items)
• Suggested follow-up actions

Limitations

• Scanned PDFs require OCR
• Complex formatting may be lost
• Images/charts cannot be fully understood