prioritize-reference-files-6e4111

Prioritize Reference Files

This skill ensures agents correctly prioritize provided reference files over external data sources, preventing the use of fabricated or incorrect information when structured data is already available.

Core Principle

When reference files are provided in task context, you MUST read and prioritize them for data before attempting web searches or generating synthetic data.

Step-by-Step Instructions

Step 1: Scan Task Context for Files

Before taking any action, identify all files provided in the task context:

• Look for file attachments, uploads, or references
• Check for common data formats: .xlsx, .csv, .json, .pdf, .docx, .txt
• Note the file names and their apparent purpose

Step 2: Read Reference Files First

Read all relevant reference files before any web searches:

# Example: Read provided Excel file
file_content = read_file(file_path="Massabama active listings.xlsx", filetype="xlsx")

# Example: Read provided CSV
file_content = read_file(file_path="data.csv", filetype="csv")

# Example: Read provided PDF
file_content = read_file(file_path="report.pdf", filetype="pdf")

Important: Handling DOCX Read Failures

If read_file(filetype='docx') fails with an error (common issue), use this fallback method:

# Method 1: Unzip and parse XML directly (docx is a zip archive)
mkdir -p temp_docx && cd temp_docx
unzip -o ../document.docx
# Extract text from word/document.xml
cat word/document.xml | grep -oP '(?<=>)[^<>]+' > extracted_text.txt

# Method 2: Use python-docx via shell
run_shell(command="python -c \"from docx import Document; doc = Document('document.docx'); print('\n'.join([p.text for p in doc.paragraphs]))\"")

# Method 3: Use shell_agent for robust extraction
extraction_result = shell_agent(task="Extract all text content from document.docx using any reliable method (unzip+XML, python-docx, or pandoc)")

After extracting via fallback:

1. Verify the extracted content is complete and readable
2. Proceed to use this extracted data as your reference file content
3. Continue with Step 3 (Extract and Validate Data) using the fallback-extracted content