08｜Claude Code Integration with Crazyrouter, Series 08: Chapter 5: Document Management and Processing

08｜Claude Code Integration with Crazyrouter Series 08: Chapter 5: Document Management and Processing#

This is the 08th post in the Crazyrouter Claude Code series. This article focuses on Claude Code Integration with Crazyrouter Series 08: Chapter 5: Document Management and Processing, with emphasis on Chapter 5: Document Management and Processing, 5.1 Document Content Extraction, and extracting key information from long documents.

Unified integration settings: Claude Code / Anthropic native clients use ANTHROPIC_BASE_URL=https://cn.crazyrouter.com; OpenAI-compatible SDKs, HTTP requests, and front-end/back-end apps use base_url=https://cn.crazyrouter.com/v1.

Guide to This Chapter#

• Who it's for: developers who are already using Claude Code, are preparing to integrate domestic models, or want to standardize team calls through Crazyrouter.
• What you'll learn: how to configure environment variables according to Crazyrouter docs, organize your workflow, and avoid /v1/v1/... issues caused by an incorrect Base URL.
• Recommended prep: create a separate API Token in the Crazyrouter Console, then follow the Claude Code integration docs to complete the base setup.

Chapter 5: Document Management and Processing#

5.1 Document Content Extraction#

Claude Code can help you quickly extract key information from long documents, greatly improving work efficiency. Whether you need a summary, key points, or table data, it can handle it with ease.

Extract Key Information from Long Documents#

Extract a Summary#

Generate a brief summary of the document:

You: Please help me generate a summary for this document (uploading a long document)

Claude: I've read the document. Here is the summary:

This document introduces the company's 2026 annual work plan. The main points include:

1. Overall goal: achieve 20% revenue growth and expand into 3 new markets
2. Market strategy: strengthen online marketing, optimize the product line, and improve customer service
3. Team building: hire 50 new employees, strengthen training, and improve team capabilities
4. Technology investment: increase R&D spending and launch 2 new products
5. Risk management: establish a risk early-warning mechanism and strengthen compliance management

Extract Key Points#

Extract the key points from the document:

You: Please extract the resolutions from this meeting memo (upload the meeting memo file)

Claude: The resolutions extracted from the meeting memo:

1. Product Development

   • Decided to develop a mobile app, expected to be completed within 6 months
   • New product features: user points system and social sharing features

2. Marketing Promotion

   • Invest 1 million yuan in online advertising
   • Reached promotional partnership agreements with 3 KOLs

3. Team Restructuring

   • Added a Marketing Manager role and hired 1 person
   • Expanded the engineering team by 5 people

4. Operations Optimization

   • Optimize the customer service process to improve response time
   • Establish a customer feedback mechanism to collect feedback regularly

Extract Specific Information#

Extract specific types of information from the document:

You: Please extract all date information from this contract (upload the contract file)

Claude: The date information extracted from the contract:

1. Contract signing date: January 15, 2024

2. Contract effective date: February 1, 2024

3. Project start date: February 15, 2024

4. Project end date: December 31, 2024

5. Payment dates:

   • Initial payment: February 1, 2024
   • Second payment: June 30, 2024
   • Final payment: January 15, 2025

6. Acceptance date: December 31, 2024

Document Summary Generation#

Brief Summary#

Generate a short summary of the document:

You: Please generate a 200-word abstract for this paper (upload the paper file)

Claude: Paper abstract:

This paper examines the application of artificial intelligence in medical diagnosis. By comparing traditional diagnostic methods with AI-assisted diagnostic methods, it finds that AI performs exceptionally well in image recognition and data analysis. The study shows that the accuracy of AI-assisted diagnosis reaches 95%, significantly higher than the 82% achieved by traditional methods. At the same time, AI can greatly shorten diagnostic time and improve healthcare efficiency. However, AI diagnosis also faces challenges such as data privacy and algorithmic bias. The article concludes with improvement suggestions, including stronger data protection, algorithm optimization, and greater transparency.

Detailed Summary#

Generate a detailed document summary:

You: Please generate a detailed summary for this report, including the background, methods, results, and conclusion (upload the report file)

Claude: Detailed summary:

Background

This report studies the company's customer churn problem. In recent years, the churn rate has risen year by year, from 5% in 2021 to 12% in 2023, severely affecting company performance.

Methods

The study uses a combination of surveys, data analysis, and in-depth interviews. A total of 1,000 valid questionnaires were collected, 2 years of customer data were analyzed, and 50 churned customers were interviewed.

Results

The study found that the main reasons for customer churn include:

1. Product prices are too high (35%)
2. Poor service quality (28%)
3. Strong attractiveness of competitors' products (22%)
4. Other reasons (15%)

Conclusion

The company is advised to take the following measures to reduce customer churn:

1. Optimize pricing strategies and offer more discount plans
2. Improve service quality and strengthen employee training
3. Increase product competitiveness and develop differentiated features
4. Establish a customer care mechanism to improve retention

Extract Table Data#

Extract Tables from a Document#

Extract table data from the document:

You: Please extract all tables from this document (upload a document containing tables)

Claude: Tables extracted from the document:

Table 1: Sales Data

Table 2: Employee Information

Analyze Table Data#

Analyze the table data:

You: Please analyze this sales data table and identify the best-selling month and product (upload the table file)

Claude: Sales data analysis:

Overall performance

• Total sales: 1,234,567 yuan
• Average monthly sales: 102,880 yuan
• Highest monthly sales: 150,000 yuan (June)
• Lowest monthly sales: 80,000 yuan (February)

Product analysis

• Best-selling product: Product A (total sales of 500,000 yuan, accounting for 40.5%)
• Fastest-growing product: Product C (grew from 30,000 yuan in January to 60,000 yuan in June, up 100%)
• Product to watch: Product B (sales declined for 3 consecutive months)

Monthly analysis

• Best-selling month: June (150,000 yuan)
• Fastest-growing month: May (30% month-over-month growth)
• Month with declining sales: February (20% month-over-month decline)

Case Example#

Case: A Teacher Organizing Teaching Materials#

Scenario: A teacher needs to extract key knowledge points from multiple teaching documents and organize them into a teaching outline.

You: Please help me extract the key knowledge points from these teaching documents and organize them into a syllabus (upload multiple teaching documents)

Claude: Syllabus organized:

Course Title: Python Programming Basics

Course Duration: 12 weeks

Target Audience: Programming beginners

Syllabus

Week 1: Introduction to Python

• History and features of Python
• Application areas of Python
• Setting up the development environment

Week 2: Basic Syntax

• Variables and data types
• Operators and expressions
• Input and output

Week 3: Control Structures

• Conditional statements (if-else)
• Loop statements (for, while)
• break and continue

Week 4: Data Structures

• Lists (List)
• Tuples (Tuple)
• Dictionaries (Dictionary)
• Sets (Set)

Week 5: Functions

• Function definition and calling
• Parameters and return values
• Scope

Week 6: Object-Oriented Programming (Part I)

• Classes and objects
• Attributes and methods
• Constructors

Week 7: Object-Oriented Programming (Part II)

• Inheritance
• Polymorphism
• Encapsulation

Week 8: Exception Handling

• try-except structure
• Common exception types
• Custom exceptions

Week 9: File Operations

• File reading and writing

• File path handling

• JSON data processing Week 10: Modules and Packages

• Importing modules

• Creating custom modules

• Common standard libraries

Week 11: Hands-on Project (Part 1)

• Project requirements analysis
• Project design
• Coding implementation

Week 12: Hands-on Project (Part 2)

• Project refinement
• Testing and debugging
• Project presentation

Tips#

1. Define the extraction goal: know exactly what you want to extract from the document, such as a summary, key points, or specific data
2. Provide formatting requirements: clearly specify the output format, such as a list, table, or paragraphs
3. Process in batches: if the document is long, extract it in batches, one section at a time
4. Verify the results: after extraction, manually review the output to ensure accuracy and completeness
5. Save the results: save the extracted content to a new document for future use
6. Build templates: for recurring extraction tasks, create templates to improve efficiency

Now, try asking Claude Code to help you extract content from some documents.

5.2 Document Format Processing#

Claude Code can help you handle a wide range of document formatting tasks: standardizing formats, converting file types, fixing issues, and making documents more consistent and professional.

Standardizing Document Formats#

Standardizing Fonts and Styles#

Make multiple documents use the same fonts and styles:

You: Please help me standardize the formatting of these documents. Use SimSun, 12pt font size, and 1.5 line spacing. (Upload multiple documents)

Claude: The document formatting has been standardized:

Standard applied

• Font: SimSun
• Font size: 12pt
• Line spacing: 1.5
• Paragraph spacing: 0.5 lines before, 0.5 lines after

Processing results

• Document 1: standardized formatting applied
• Document 2: standardized formatting applied
• Document 3: standardized formatting applied

All documents now use a consistent format and look more professional and standardized.

Standardizing Heading Formats#

Make the headings in a document use a consistent format:

You: Please help me standardize the heading formats in this document. Use SimHei 16pt for level-one headings and SimHei 14pt for level-two headings. (Upload document)

Claude: The heading formats have been standardized:

Heading format standards

• Level-one heading: SimHei, 16pt, centered
• Level-two heading: SimHei, 14pt, left-aligned
• Level-three heading: SimHei, 12pt, left-aligned

Processing result

All headings in the document have been formatted according to the standards above, making the document structure clearer.

Standardizing Numbering Formats#

Make the numbering in a document use a consistent format:

You: Please help me standardize the numbering format in this document. Use "I, II, III" as level-one numbering and "(I) (II) (III)" as level-two numbering. (Upload document)

Claude: The numbering format has been standardized:

Numbering format standards

• Level-one numbering: I, II, III, IV...
• Level-two numbering: (I) (II) (III) (IV)...
• Level-three numbering: 1. 2. 3. 4...

Processing result

All numbering in the document has been formatted according to the standards above, making the document hierarchy clearer.

Converting Document Types#

Word to PDF#

Convert a Word document to PDF:

You: Please help me convert this Word document to PDF. (Upload Word document)

Claude: The Word document has been converted to PDF:

Conversion result

• Original file: report.docx
• Converted file: report.pdf
• File size: 2.5 MB
• Pages: 15

The PDF document has been generated, with formatting and content consistent with the original Word document.

PDF to Word#

Convert a PDF document to Word:

You: Please help me convert this PDF document to Word. (Upload PDF document)

Claude: The PDF document has been converted to Word:

Conversion result

• Original file: manual.pdf
• Converted file: manual.docx
• File size: 1.8 MB
• Pages: 20

Notes

• The original formatting has been preserved as much as possible, but minor adjustments may still be needed
• Images have been extracted and embedded in the Word document
• Tables have been converted to Word table format

Text to Markdown#

Convert plain text to Markdown:

You: Please help me convert this text file to Markdown. (Upload text file)

Claude: The text file has been converted to Markdown:

# Document Title

## Chapter 1
This is the content of Chapter 1...

## Chapter 2
This is the content of Chapter 2...

### 2.1 Section
This is the content of the section...

### 2.2 Section
This is the content of the section...

The Markdown document has been generated, with a clear format that is easy to read and edit.

Fixing Document Issues#

Fixing Formatting Errors#

Fix formatting errors in a document:

You: Please help me fix the formatting errors in this document. Some paragraphs use inconsistent formatting. (Upload document)

Claude: The document formatting errors have been fixed:

Issues found

1. Paragraph 3 used a different font
2. Paragraph 5 had inconsistent line spacing
3. Paragraph 7 had incorrect indentation
4. Paragraph 10 used an inconsistent font size

Fixes applied

• Standardized the font for all paragraphs
• Standardized the line spacing for all paragraphs
• Corrected the indentation for all paragraphs
• Standardized the font size for all paragraphs

The document formatting is now consistent and standardized.

Fixing Numbering Errors#

Fix numbering errors in a document:

You: Please help me fix the numbering errors in this document. The numbering is a bit messy. (Upload document)

Claude: The document numbering errors have been fixed:

Issues found

1. Chapter 3 had the wrong number (it should have been III, but was written as IV)
2. Chapter 5 was missing a number
3. Chapter 7 had duplicate level-two numbering
4. Chapter 9 had numbering in the wrong order

Fixes applied

• Corrected the numbering for Chapter 3
• Added the missing number for Chapter 5
• Corrected the level-two numbering in Chapter 7
• Corrected the numbering order in Chapter 9

The document numbering is now correct and continuous.

Fixing Image Issues#

Fix image issues in a document:

You: Please help me fix the image issues in this document. Some images are not displaying correctly. (Upload document)

Claude: The document image issues have been fixed:

Issues found

1. The image on page 3 had low resolution
2. The image on page 5 was misaligned
3. The image on page 7 had an inappropriate size
4. The image on page 9 was missing a caption

Fixes applied

• Increased the resolution of the image on page 3
• Adjusted the position of the image on page 5
• Adjusted the size of the image on page 7
• Added a caption to the image on page 9

The images in the document now display correctly.

Example Use Cases#

Example: A Designer Organizes Design Guidelines#

Scenario: A designer needs to organize multiple design documents, standardize their formatting, and produce a design guidelines document.

You: Please help me organize these design documents, standardize the formatting, and create a design guidelines document. (Upload multiple design documents)

Claude: The design guidelines document has been organized:

Design Guidelines Document

1. General Principles

1.1 Design Principles

• Simplicity: the design should be clear and concise, avoiding redundancy
• Consistency: maintain a consistent design style
• Usability: the design should be easy to use and understand
• Aesthetics: the design should look polished and appealing

1.2 Scope

• These guidelines apply to all product design work in the company
• This includes web pages, apps, promotional materials, and more

1. Color Guidelines

2.1 Primary Colors

• Brand color: #3B82F6 (blue)
• Secondary color: #10B981 (green)
• Alert color: #EF4444 (red)

2.2 Neutral Colors

• Black: #000000
• Dark gray: #374151
• Medium gray: #9CA3AF
• Light gray: #E5E7EB
• White: #FFFFFF

2.3 Usage Rules

• Use the primary color for primary buttons, links, and similar elements
• Use the secondary color for success states, confirmation actions, and similar elements
• Use the alert color for error messages, warning messages, and similar elements

1. Typography Guidelines

3.1 Font Selection

• Chinese: Source Han Sans
• English: Roboto
• Numbers: Roboto Mono

3.2 Font Size Guidelines

• Level-one heading: 24pt
• Level-two heading: 20pt
• Level-three heading: 16pt
• Body text: 14pt
• Supporting text: 12pt

3.3 Usage Rules

• Use bold for headings
• Use regular weight for body text
• Use light weight for supporting text

1. Icon Guidelines

4.1 Icon Style

• Line icons
• Consistent stroke width: 2px
• Consistent border radius: 4px

4.2 Icon Sizes

• Small icon: 16x16px
• Medium icon: 24x24px
• Large icon: 32x32px

4.3 Usage Rules

• Icons should be simple and clear
• Avoid excessive detail
• Keep the style consistent

1. Spacing Guidelines

5.1 Base Spacing

• Minimum spacing: 4px
• Small spacing: 8px
• Medium spacing: 16px
• Large spacing: 24px
• Extra-large spacing: 32px

5.2 Usage Rules

• Use medium spacing between elements
• Use large spacing between sections
• Use extra-large spacing for page margins

Tips#

1. Create formatting templates: for commonly used document formats, create templates to improve efficiency
2. Batch processing: upload multiple documents at once to batch-process formatting issues
3. Back up original files: before modifying documents, it is recommended to back up the originals
4. Verify step by step: after processing, review the documents step by step to ensure the formatting is correct
5. Save the guidelines: save the finalized formatting guidelines for future use
6. Update regularly: as requirements change, update the formatting guidelines on a regular basis Now, try using Claude Code to help you handle some document formatting tasks!

5.3 Batch Document Processing#

Claude Code can help you process multiple documents in batches, significantly improving your productivity. Whether you need to batch rename files, batch edit content, or batch generate documents, it can handle the work easily.

Batch Renaming#

Rename by Rules#

Batch rename files according to specific rules:

You: Please help me batch rename these files using this format: date_title_sequence.extension (upload multiple files)

Claude: The files have been batch renamed:

Renaming rules

• Format: date_title_sequence.extension
• Date: YYYY-MM-DD format
• Title: extracted from the file content
• Sequence: increments starting from 001

Renaming results

• report1.docx → 2024-01-15_Sales_Report_001.docx
• report2.docx → 2024-01-16_Market_Analysis_002.docx
• report3.docx → 2024-01-17_Product_Planning_003.docx
• report4.docx → 2024-01-18_Financial_Statement_004.docx

All files have been renamed according to the rules.

Rename by Content#

Batch rename files based on their content:

You: Please help me batch rename these files based on the first-line title in each file (upload multiple files)

Claude: The files have been batch renamed based on their content:

Renaming rules

• Use the first line of each file as the file name
• Remove special characters
• Keep the original extension

Renaming results

• file1.txt → Introduction_to_Python_Programming.txt
• file2.txt → Data_Analysis_Fundamentals.txt
• file3.txt → Practical_Machine_Learning.txt
• file4.txt → Advanced_Deep_Learning.txt

All files have been renamed based on their content.

Rename by Date#

Batch rename files by file date:

You: Please help me batch rename these files by their creation date (upload multiple files)

Claude: The files have been batch renamed by creation date:

Renaming rules

• Format: YYYY-MM-DD_original_filename.extension
• Date: file creation date

Renaming results

• document1.pdf → 2024-01-10_document1.pdf
• document2.pdf → 2024-01-12_document2.pdf
• document3.pdf → 2024-01-15_document3.pdf
• document4.pdf → 2024-01-18_document4.pdf

All files have been renamed by creation date.

Batch Content Editing#

Batch Text Replacement#

Batch replace text across multiple files:

You: Please help me replace all occurrences of "Product A" with "New Product X" in these files (upload multiple files)

Claude: Text replacement has been completed in all files:

Replacement details

• Find text: Product A
• Replacement text: New Product X
• Files processed: 5
• Total replacements: 23

Replacement count by file

• File 1: 5 replacements
• File 2: 4 replacements
• File 3: 6 replacements
• File 4: 4 replacements
• File 5: 4 replacements

All occurrences of "Product A" in the files have been replaced with "New Product X".

Batch Format Changes#

Batch change formatting across multiple files:

You: Please help me change the font of all headings in these files to SimHei and the font size to Size 3 (upload multiple files)

Claude: Formatting changes have been completed in all files:

Change details

• Target: all headings
• Font: changed to SimHei
• Font size: changed to Size 3 (16pt)
• Files processed: 8

Changes by file

• File 1: changed 3 headings
• File 2: changed 4 headings
• File 3: changed 3 headings
• File 4: changed 5 headings
• File 5: changed 4 headings
• File 6: changed 3 headings
• File 7: changed 4 headings
• File 8: changed 3 headings

The heading format has been standardized across all files.

Batch Add Content#

Batch add content to multiple files:

You: Please help me add the company copyright statement to the beginning of each of these files (upload multiple files)

Claude: The copyright statement has been added to the beginning of all files:

Added content

Copyright © 2024 XX Company
This document may not be reproduced or used without permission

Processing results

• Files processed: 10
• The copyright statement has been added to the beginning of all files

The copyright statement has been added to all files.

Batch Document Generation#

Batch Generate Reports#

Batch generate reports from data:

You: Please help me generate a monthly report for each department based on the data in this Excel spreadsheet (upload Excel file)

Claude: Monthly reports have been generated for each department:

Generated report list

1. Sales_Department_Monthly_Report.docx
2. Marketing_Department_Monthly_Report.docx
3. Engineering_Department_Monthly_Report.docx
4. Finance_Department_Monthly_Report.docx
5. HR_Department_Monthly_Report.docx

Report content

Each report includes:

• Department overview
• Monthly work summary
• Data analysis
• Issues and recommendations
• Plan for next month

All reports have been generated and saved.

Batch Generate Notices#

Batch generate notices from a template:

You: Please help me generate onboarding notices for these employees based on this template (upload the template and employee list)

Claude: Onboarding notices have been generated for all employees:

Generated notice list

1. Alice_Onboarding_Notice.docx
2. Bob_Onboarding_Notice.docx
3. Charlie_Onboarding_Notice.docx
4. David_Onboarding_Notice.docx
5. Emily_Onboarding_Notice.docx

Notice content

Each notice includes:

• Employee name
• Start date
• Department
• Position
• Check-in time
• Check-in location
• Required materials

All notices have been generated and saved.

Batch Generate Emails#

Batch generate emails from a template:

You: Please help me generate thank-you emails for these customers based on this template (upload the template and customer list)

Claude: Thank-you emails have been generated for all customers:

Generated email list

1. Customer_A_Thank_You_Email.txt
2. Customer_B_Thank_You_Email.txt
3. Customer_C_Thank_You_Email.txt
4. Customer_D_Thank_You_Email.txt
5. Customer_E_Thank_You_Email.txt

Email content

Each email includes:

• Customer name
• Thank-you message
• Collaboration recap
• Future outlook
• Contact entry point

All emails have been generated and saved.

Case Example#

Case: Finance Staff Processing Invoice Records#

Scenario: A finance staff member needs to process a large number of invoice records, including batch renaming, data extraction, and report generation.

You: Please help me process these invoice files:

1. Batch rename them as: invoice_number_date_amount.pdf
2. Extract the key information from all invoices
3. Generate a summary report (upload multiple invoice files)

Claude: Batch invoice processing is complete:

1. Batch rename

All invoice files have been renamed:

• INV001_2024-01-05_1200.pdf
• INV002_2024-01-08_3500.pdf
• INV003_2024-01-12_2800.pdf
• INV004_2024-01-15_4200.pdf
• INV005_2024-01-18_5600.pdf

1. Extract key information

1. Generate summary report

January 2024 Invoice Summary Report

1. Overall Summary#

• Number of invoices: 5
• Total invoice amount: 17,300 yuan
• Total tax amount: 2,249 yuan
• Amount excluding tax: 15,051 yuan

2. Statistics by Supplier#

3. Statistics by Date#

IV. Analysis Conclusions#

1. This month’s invoice amounts were mainly concentrated in the latter half of the month (15th–18th), accounting for 56.6% of the total
2. Supplier E was the largest supplier this month, accounting for 32.4%
3. The average invoice amount was 3,460 yuan

Tips#

1. Test before batching: Before batch processing, test on a single file first and proceed in bulk only after confirming the result
2. Back up original files: Before batch processing, always back up the original files in case something goes wrong
3. Use wildcards: When batch renaming, you can use wildcards (*, ?) to match filenames
4. Keep processing logs: After batch processing, save the processing logs for later review and traceability
5. Process in batches: If there are many files, process them in smaller batches to avoid handling too many at once
6. Verify the results: After batch processing, randomly check a few files to confirm that the results are correct

Now, try asking Claude Code to help you batch process some documents!

Related Reading#

• Previous article: Part 07
• Next article: Part 09
• Configuration docs: Connect Claude Code to Crazyrouter
• API endpoint guide: Base URL and /v1 usage
• One-click configuration script: Crazyrouter Claude Code One-Click Configuration Script

Start Integrating Crazyrouter#

If you’re ready to connect Claude Code, Chinese models, or your own applications to Crazyrouter through a unified setup, follow this sequence:

1. Go to the Crazyrouter Console, create a dedicated API Token, and manage permissions separately by project or team.
2. Claude Code should use the root domain: https://cn.crazyrouter.com; OpenAI-compatible SDKs should use: https://cn.crazyrouter.com/v1.
3. When you need to automatically check the environment or quickly write configuration, use the Crazyrouter Claude Code One-Click Configuration Script.
4. If debugging fails, check the console logs first, then review the API Endpoint guide. Pay special attention to whether the Base URL has an extra /v1.

When you need to evaluate model costs or choose between different models, check the Crazyrouter pricing and models page first, then add your commonly used models to the Token whitelist.