Module team_billing 

GET /v1/storage + GET /v1/usage — the team server’s billing-plane surface (docs/contracts/billing-plane-v2.md, GL #463).

/v1/storage reports the hosted workspace footprint (retrieval index, knowledge store, event log — everything the server persists under its data root and the workspaces’ .lean-ctx state dirs). It is server-measured: the control plane’s hourly metering_job polls it for Stripe meter events and threshold mails, so the report carries plain numbers and no content. Field casing is camelCase (usedBytes), matching what lean-ctx-cloud/src/metering_job.rs and metering.rs::from_storage read.

/v1/usage is the unified usage snapshot for the account dashboard: the savings roll-up (from the same signed-batch store as /v1/savings/summary) plus a storage block in snake_case (used_bytes) — the spelling metering.rs::from_usage expects for that block.

Sizing uses allocated disk blocks (st_blocks * 512) on Unix so sparse and partially-written files bill what they actually occupy; on other platforms it falls back to logical file length. Reports are cached for STORAGE_CACHE_TTL per process — the walk is O(files) and the metering job polls hourly, so 60 s keeps repeated dashboard hits cheap without letting bills go stale.

Authorisation: both routes are gated by TeamScope::Audit in the team auth middleware — same sensitivity class as /v1/metrics and /v1/savings/summary, and the scope the control plane’s audit-only token carries.

Structs

StorageCache

Cached report + when it was measured.

StorageComponent

One measured storage component (a directory or file the server persists).

StorageReport

A measured (uncached) storage report.

StorageRoots

The measurement inputs, fixed at server startup.

Functions

v1_storage

GET /v1/storage — camelCase, served from the 60 s cache.

v1_usage

GET /v1/usage — savings roll-up + snake_case storage block.