For AI agents: fetch the complete documentation index at /llms.txt, the full documentation corpus at /llms-full.txt, the agent skill at /skill.md, or this page as Markdown at /docs/mobile/usage.md (or send Accept: text/markdown).
Usage and credits
Metered limits, prepaid credit, and the billing surface.
Read the usage state
GET /api/mobile/usage returns the AI budget as microunit decimal strings — budgetMicrounits, committedMicrounits, heldMicrounits, and keyMode — the prepaid credit balance, per-meter rows with label, used, limit, cap, planLimit and description, and the current period bounds. Presentation marks a meter warning at eighty percent and danger at one hundred.
Values are decimal strings, not JSON numbers — parse them before comparing or formatting so large budgets stay exact.
Prepaid credit ledger
GET /api/mobile/credits returns the workspace's prepaid balance and its last fifty ledger rows. The balance feeds Settings → Billing and the home usage card; ledger reference ids are withheld from the client payload.
Home and billing surfaces
The home screen reuses GET /api/workspace/metrics — the same stat grid the web dashboard renders: monthly AI spend, task and thread counts, members, pending invitations, plan status, and unread notifications. Its usage card shows the AI row plus finite-limit meters, and quick-action tiles link to chat, assistant, content, commerce, notifications, feedback, and settings.
Settings → Billing shows the plan, expiration, credit balance, and usage meters, explains store-versus-web billing, and links to the paywall. A server schedule sweeps workspaces every fifteen minutes and notifies owners and admins once a month at eighty and one hundred percent of the AI budget.
Verification and troubleshooting
Check a meter near its cap, an exhausted credit balance, and a workspace on the default budget. A blocked action must point at the meter that denied it, not at a generic error.