Provider usage
Provider Usage surfaces whatever capacity or quota signal a provider CLI actually reports, rather than inferring a number it can't verify.
Provider Usage gives you visibility into rate limits and capacity as reported by each agent's own CLI — refreshed manually, and only ever showing values AgentDock can actually parse.
What it shows
Each provider reports capacity information differently, so what you see depends on the agent:
- An exact usage or quota value, when the provider CLI exposes one that AgentDock can parse.
- A general capacity/rate-limit signal (for example, a detected cooldown or reset time) when an exact figure isn't available but a capacity condition is still detected.
Note
If a provider doesn't expose a parseable quota value, AgentDock shows that usage information is unavailable rather than guessing a percentage. Don't treat a missing number as zero usage.
How this connects to capacity errors
The same underlying capacity signal Provider Usage displays is what AgentDock (and, in AgentDock Pro, AIgency's capacity recovery) uses to recognize a rate-limit or quota error as a temporary condition rather than a genuine task failure. See Capacity or rate limits if a session appears blocked on this.
Related
- Provider Usage in AIgency — the same signal, used for automatic recovery in orchestrated runs.
- Capacity or rate limits