docs/assets/example-scripts/billing

Actions Cost Summary Script

The script can be found here: https://gist.github.com/docs-bot/98cb03ec43b716b1f8e03bcc091d069c.

This script processes GitHub Actions billing data to generate cost summaries by SKU (Stock Keeping Unit).

This script is referenced in the Billing tutorial: content/billing/tutorials/estimate-actions-costs.md

Purpose

The script analyzes summarized usage reports from GitHub billing and creates CSV summaries that help you understand:

• Which Actions SKUs are being used
• Total costs per SKU
• Free minutes usage per SKU

Optionally, the script can generate per-organization breakdowns to analyze usage across different organizations.

Requirements

• Python 3.x (uses only standard library modules)

Use the script

Basic usage

Generate a single SKU summary file:

python3 summarize_actions_costs.py input_file.csv

This creates input_file_sku.csv containing:

• Each distinct SKU that starts with actions_
• Sum of all costs (net_amount) for each SKU
• Sum of free minutes used (where net_amount is 0) for each SKU

Per-organization summaries

Generate both the SKU summary and individual files for each organization:

python3 summarize_actions_costs.py input_file.csv --by-org

This creates:

• input_file_sku.csv - Overall SKU summary across all organizations
• input_file.organization_name.csv - Individual summary for each organization

Each organization file contains the same SKU-level breakdown (costs and free minutes) but filtered to that specific organization's usage.

Output format

All output CSV files have the following format:

sku,total_net_amount,free_minutes_quantity
actions_linux,10800.30,248594.00
actions_macos,24424.38,10294.00
...

• sku: The SKU identifier (only those starting with actions_)
• total_net_amount: Sum of all costs for this SKU
• free_minutes_quantity: Sum of quantity used when no cost was incurred (free tier usage)

Notes

• The script only processes rows where the product is actions
• Only SKUs starting with actions_ are included in the output
• Organization names are sanitized for safe filenames (removes special characters, path separators, and limits length)
• Empty organization names are saved as unknown
• The script validates required CSV columns before processing
• Malformed rows are skipped with a warning message
• CSV injection protection is applied to output fields
• Maximum of 1000 organizations when using --by-org to prevent resource exhaustion