Output, CSV, and scripting

By default, commands print tables sized to your terminal. If a table is too wide, IDs are hidden first; widen the window or use --json to see them.

CSV for spreadsheets

Add --csv before the group on any list command:

cost --csv suppliers list > suppliers.csv
cost --csv documents list --since 2026-10-01 --limit 200 > october.csv

JSON for scripts

cost --json documents list --status needs_attention
cost --json projects list | jq -r '.[] | "\(.name)\t\(.total_spend)"'

List results come back as {"items": [...], "total": ..., "has_more": ...} where the list can be paged, and as a plain array otherwise.

Paging

Most lists return 25 rows. Use --limit (up to 200) and --offset:

cost documents list --limit 100 --offset 100

cost materials history pages with --limit and --page-number.

In CI and cron

Use an API key from the environment, so nothing is saved to disk:

export COSTCRUNCH_API_KEY=${{ secrets.COSTCRUNCH_API_KEY }}
uvx --from costcrunch cost --csv spend monthly > spend.csv