file-generation

name	file-generation
description	Generate files (charts, documents, spreadsheets, images, presentations) by executing Python code in a sandboxed environment
license	Apache-2.0
compatibility	>=1.0

File Generation

When users request file generation (charts, documents, spreadsheets, images, presentations), you have access to the generate_file tool that can execute Python code to create these files.

CRITICAL INSTRUCTIONS FOR FILE GENERATION:

• NEVER mention file paths, directories, or sandbox links in your response
• NEVER include details like /tmp/muxi_artifacts/, file sizes, or download links
• DO NOT provide any technical file system information to the user
• Simply acknowledge what you created (e.g., "I've created the chart you requested")
• The files are automatically attached as artifacts - you don't need to explain this
• Focus on describing WHAT you created, not WHERE it was saved

How to use the file generation tool:

1. Understand the user's request - Identify what type of file they need (chart, document, spreadsheet, etc.)

2. Write complete Python code that:

   • Imports only from allowed libraries
   • Creates the desired output
   • Saves the file with an appropriate name and extension
   • Does not attempt to access files outside the working directory

3. Use the generate_file tool with your code:

   generate_file(code="your_python_code_here", filename="optional_hint.ext")
   

Available libraries for file generation:

Data Processing & Analysis:

• pandas - Data manipulation and analysis
• numpy - Numerical computing
• scipy - Scientific computing
• statsmodels - Statistical modeling

Visualization & Charts:

• matplotlib - 2D plotting library (always use matplotlib.use('Agg'))
• seaborn - Statistical data visualization
• plotly - Interactive visualizations
• bokeh - Interactive visualization library
• altair - Declarative visualization

Document Generation:

• python-docx - Create/modify Word documents (.docx)
• reportlab - PDF generation (prefer over fpdf for Unicode support)
• fpdf2 - Simple PDF generation
• markdown - Markdown processing

Spreadsheets:

• openpyxl - Read/write Excel 2010 xlsx/xlsm files
• xlsxwriter - Create Excel files
• pandas with to_excel() - Excel export

Images & Graphics:

• PIL/Pillow - Image processing
• qrcode - QR code generation
• python-barcode - Barcode generation

Presentations:

• python-pptx - Create PowerPoint presentations

Other Formats:

• pyyaml - YAML files
• lxml - XML processing
• json - JSON files
• csv - CSV files

Networking:

• requests - HTTP requests
• urllib - URL handling

Example patterns:

Creating a chart:

import matplotlib
matplotlib.use('Agg')
import matplotlib.pyplot as plt
import numpy as np

x = np.linspace(0, 10, 100)
y = np.sin(x)

plt.figure(figsize=(10, 6))
plt.plot(x, y, 'b-', linewidth=2)
plt.title('Sine Wave')
plt.xlabel('X values')
plt.ylabel('Y values')
plt.grid(True, alpha=0.3)
plt.savefig('sine_wave.png', dpi=300, bbox_inches='tight')
plt.close()

Creating a document:

from docx import Document
from docx.shared import Inches, Pt

doc = Document()
doc.add_heading('Report Title', 0)

p = doc.add_paragraph('This is an example paragraph with ')
p.add_run('bold text').bold = True
p.add_run(' and ')
p.add_run('italic text').italic = True

doc.add_paragraph('Key findings:', style='List Bullet')
doc.add_paragraph('First finding', style='List Bullet')
doc.add_paragraph('Second finding', style='List Bullet')

doc.save('report.docx')

Creating a spreadsheet:

import pandas as pd
import numpy as np

data = {
    'Date': pd.date_range('2024-01-01', periods=10),
    'Sales': np.random.randint(100, 1000, 10),
    'Costs': np.random.randint(50, 500, 10)
}

df = pd.DataFrame(data)
df['Profit'] = df['Sales'] - df['Costs']

with pd.ExcelWriter('sales_report.xlsx', engine='openpyxl') as writer:
    df.to_excel(writer, sheet_name='Sales Data', index=False)

Creating a QR code:

import qrcode

qr = qrcode.QRCode(version=1, error_correction=qrcode.constants.ERROR_CORRECT_L, box_size=10, border=4)
qr.add_data('https://example.com')
qr.make(fit=True)
img = qr.make_image(fill_color="black", back_color="white")
img.save('qr_code.png')

Creating a presentation:

from pptx import Presentation
from pptx.util import Inches, Pt

prs = Presentation()

title_slide_layout = prs.slide_layouts[0]
slide = prs.slides.add_slide(title_slide_layout)
slide.shapes.title.text = "Quarterly Report"
slide.placeholders[1].text = "Q4 2024 Results"

content_slide_layout = prs.slide_layouts[1]
slide = prs.slides.add_slide(content_slide_layout)
slide.shapes.title.text = "Key Metrics"
slide.placeholders[1].text = "- Revenue: $1.2M\n- Growth: 25%\n- New Customers: 150"

prs.save('presentation.pptx')

Important constraints:

1. Always save files - Your code must explicitly save the output file
2. Use descriptive filenames - Include appropriate extensions (.png, .xlsx, .docx, etc.)
3. Handle errors gracefully - Consider what might go wrong and handle it
4. Keep code focused - Generate only what the user requested
5. Avoid external dependencies - Use only the allowed libraries listed above
6. Do not access external resources - No file system access outside the working directory
7. Prefer reportlab over fpdf - fpdf's default font cannot encode non-ASCII characters
8. Always set matplotlib backend - Use matplotlib.use('Agg') before importing pyplot

Response Examples:

GOOD Response: "I've created the bar chart showing your Q1 sales data with the three categories you requested."

BAD Response: "I've created the chart at /tmp/muxi_artifacts/chart.png (15KB). You can download it here: [link]"

Remember: Keep responses simple and focused on what was created, not technical details.