Find your top credit burners

gh-aw-fleet consumption rolls up AI-credit (AIC) usage across the fleet from gh aw logs. Change the --by axis and the window to answer different “where did the credits go?” questions. It is read-only and needs no deployed reporting workflow.

Rank repositories

Start with the per-repo rollup (the default axis):

gh-aw-fleet consumption --by repo

Each row is a repository with its API calls, safe writes, AIC, and derived USD (AIC × $0.01). The highest AIC rows are your biggest spenders.

Rank individual workflows

To find the single workflows burning the most, group by workflow. The output includes a TOP 10 BURNERS footer ranked by AIC:

gh-aw-fleet consumption --by workflow

Drill into one repository

Scope the rollup to one repo and break it down by workflow to see where that repo’s spend concentrates:

gh-aw-fleet consumption you/your-repo --by workflow

Narrow the time window

consumption defaults to the most-recent data per repo (--latest). To measure a period instead, use a trailing window or a start date (the three are mutually exclusive):

gh-aw-fleet consumption --trailing 7d --by workflow    # last 7 days
gh-aw-fleet consumption --since 2026-06-01 --by repo   # since a date

Flag rows over a ceiling

Add --budget <AIC> to mark rows whose AIC strictly exceeds a ceiling with a ! in a trailing OVER column — useful for spotting concentration at a glance:

gh-aw-fleet consumption --by repo --budget 500

Export for a cost system

Emit the JSON envelope to feed the numbers into a spreadsheet or cost tool:

gh-aw-fleet consumption --by cost-center --output json