Capacity recovery
AIgency rechecks real provider availability on a bounded backoff schedule instead of retrying blindly or waiting on a displayed reset time alone.
AgentDock Pro
Capacity recovery is part of AIgency, available only in AgentDock Pro.
A task blocked on a provider rate limit or quota moves to waiting_capacity rather than being treated as a failure — see Execution. Capacity recovery is what gets it out of that state without you having to babysit it.
Known vs. unknown reset time
- Known reset time — when a provider reports when capacity will return, AIgency uses that as a target for when to recheck.
- Unknown reset time — without one, AIgency rechecks on a bounded backoff schedule instead of polling indefinitely: roughly 2 minutes, then 5, then 10, then 15 minutes between checks.
Note
A reported reset time is a signal, not a guarantee — AIgency verifies real provider availability before resuming work rather than trusting the clock alone.
What resumes the task
A waiting_capacity task returns to pending when you choose "Continue when reset" or "Try again now," when the reset time reconciles automatically, or when you switch the task to a different agent. Time spent waiting never counts against the task's retry budget, and this wait state survives an app restart — see Restart and crash recovery.