analytics

Library Analytics: who inserts and detaches your published library assets, and where those assets are in use. The file argument is always the published library file, by key or URL.

Enterprise plan

Library Analytics needs an Enterprise plan and the library_analytics:read scope. Unlike the other Enterprise domains here, it works with a personal access token, a plan token, or OAuth.

Six commands, one per (asset, metric) pair:

Command	--group-by	Date range
analytics component-actions <file>	component | team	yes
analytics component-usages <file>	component | file	no
analytics style-actions <file>	style | team	yes
analytics style-usages <file>	style | file	no
analytics variable-actions <file>	variable | team	yes
analytics variable-usages <file>	variable | file	no

--group-by is required on all six, and only accepts the two values in its row — that is Figma’s constraint, and the CLI rejects anything else before spending the call.

-actions: a weekly time series

cyber-figma analytics component-actions <file> --group-by component

Insertions and detachments, week by week.

Option	Description
--group-by <dimension>	Required. component/style/variable, or team
--start-date <YYYY-MM-DD>	Earliest week to include — rounded back to a week start
--end-date <YYYY-MM-DD>	Latest week to include — rounded forward to a week end
--cursor <cursor>	Where to resume
--all / --max-pages <number>	Fetch every page, up to a ceiling (default: 10)

cyber-figma analytics component-actions abc123 --group-by team \
  --start-date 2026-01-01 --end-date 2026-03-31
cyber-figma analytics style-actions abc123 --group-by style --all --json

Dates are whole weeks, not days: whatever you pass is widened outward to week boundaries. With no range at all, Figma returns the prior year up to the latest computed week.

-usages: a snapshot

cyber-figma analytics component-usages <file> --group-by component

Where assets are in use right now. These take no date range at all — there is no time dimension to filter — so --start-date and --end-date are not offered, and passing one is an unknown-flag usage error that exits 2.

Option	Description
--group-by <dimension>	Required. component/style/variable, or file
--cursor <cursor>	Where to resume
--all / --max-pages <number>	Fetch every page, up to a ceiling (default: 10)

Reading the results

Row shape differs per grouping dimension, so the table columns come from the rows Figma actually sent rather than a fixed list. Text mode shows the first 60 rows and says how many it withheld; --full, --json, or --toon give you all of them.

Three semantics worth knowing before you build on this data:

• Recomputed daily at 00:00 UTC. Polling more often gains nothing.
• Rows you lack permission for are obfuscated, not dropped — they come back named Team not visible or File not visible. They are many different entities wearing one name, so aggregating them as a single team or file is wrong.
• Usage outside your organization is excluded entirely, so these numbers are an internal adoption measure, not a total.