Tax Doc Combiner

1. Find invoices matching pattern YYYY house ####-Street invoice*.pdf
2. Read each PDF to extract vendor, amount, date, and service description
3. Auto-categorize each expense based on content
4. Generate summary table as first page
5. Combine PDFs with summary first, then all invoices
6. Move originals to processed/ subfolder

Expense Categories

Summary Table Format

The first page of the master invoice contains:

╔══════════════════════════════════════════════════════════════════════════╗
║                    PROPERTY EXPENSE SUMMARY                               ║
║                    YYYY - ####-Street                                     ║
╠══════════════════════════════════════════════════════════════════════════╣
║ Date       │ Vendor                │ Category    │ Description  │ Amount ║
╠════════════╪═══════════════════════╪═════════════╪══════════════╪════════╣
║ 2024-03-15 │ ABC Plumbing          │ plumbing    │ Water heater │ $450.00║
║ 2024-05-20 │ Green Lawn Services   │ yard        │ Spring maint │ $275.00║
║ 2024-08-10 │ Cool Air HVAC         │ hvac        │ AC repair    │ $850.00║
╠════════════╧═══════════════════════╧═════════════╧══════════════╧════════╣
║                                              TOTAL:            │$1,575.00║
╠══════════════════════════════════════════════════════════════════════════╣
║ CATEGORY TOTALS                                                          ║
╠═════════════════╪════════════════════════════════════════════════════════╣
║ plumbing        │ $450.00                                                ║
║ yard            │ $275.00                                                ║
║ hvac            │ $850.00                                                ║
╚═════════════════╧════════════════════════════════════════════════════════╝

Instructions

Step 1: Find invoices for a property

ls "<YEAR>/"*invoice*.pdf 2>/dev/null | grep -i "house"

Or for a specific property:

ls "<YEAR>/"*"house"*"<PROPERTY>"*"invoice"*.pdf

Step 2: Read each invoice PDF

For each invoice, extract:

• Date: Invoice date or service date
• Vendor: Company name
• Amount: Total amount (look for "Total", "Amount Due", "Balance")
• Description: Service performed

Step 3: Categorize expenses

Match invoice content against category keywords table above. Use general as fallback.

Step 4: Generate summary and combine

Run the combiner script:

python3 .claude/skills/tax-doc-combiner/combine_invoices.py \
    --year YYYY \
    --property "####-Street" \
    --input-dir "<YEAR>/" \
    --output "<YEAR>/YYYY houses ####-Street master-invoice.pdf"

Step 5: Move processed files

mkdir -p "<YEAR>/processed"
mv "<YEAR>/YYYY house ####-Street invoice"*.pdf "<YEAR>/processed/"

Step 6: Report results

List:

• Number of invoices combined
• Total expenses by category
• Grand total
• Output file location

Dependencies

Install required Python packages:

make -C .claude/skills/tax-doc-combiner install

Notes

• Always read PDFs before processing to verify they are invoices
• If amount cannot be extracted, ask user for the total
• If category is ambiguous, ask user to confirm
• Keep original files in processed/ folder (don't delete)
• Master invoice filename uses plural "houses" to distinguish from source invoices

Future Document Types

Additional document type workflows will be added here as needed.