Committing as "net: exponential backoff for retries".
Proof on every change
Trace is the project’s memory: a timeline of sealed session records where every commit is tied to the goal it delivered, with a verdict you can hand to anyone, and they can check it without trusting you.
A verdict on the goal
Every change is checked against the goal it was meant to deliver, and the result is sealed in: “proven 3 of 3”, not a vague green check.
Make network retries back off exponentially instead of pausing a flat 500 ms every time: cap the wait at 30 s so a rate-limited service gets progressively more room to recover.
Aura locked exactly what the AI changed and why. Nothing in this record has been altered since.
Files changed1
What you asked here
Make network retries back off exponentially instead of pausing a flat 500 ms every time: cap the wait at 30 s so a rate-limited service gets progressively more room to recover.
Details
The whole story, replayable
Walk the project from the first commit: the intent, the agent session, and the exact code that shipped (who, which agent, and why).
Make network retries back off exponentially instead of pausing a flat 500 ms every time: cap the wait at 30 s so a rate-limited service gets progressively more room to recover.
Committed b7e90d4. Retries now back off exponentially instead of pausing a flat half-second.
Running the retry suite.
All four retry tests pass: the backoff grows as expected and the max-attempts guard still holds.
Let me look at how retries work today before I change anything.
Every retry sleeps a flat 500 ms via sleep_fixed. I'll replace that with an exponential backoff that grows after each failed attempt, and take the operation by FnMut so stateful closures can be retried.
Done. compute_delay doubles the wait each attempt and retry_with_backoff now sleeps for it instead of the flat delay.
Proof you can hand over
Records are cryptographically sealed and portable. Send one to a teammate, an auditor or a customer and they can verify it without you in the room.
Make network retries back off exponentially instead of pausing a flat 500 ms every time: cap the wait at 30 s so a rate-limited service gets progressively more room to recover.
What was sealed
Public witness
Every change the AI makes here is sealed the moment it happens, so you always have a true record of what happened that can't be quietly altered later. Aura checks the seal right on your own computer. It works even offline. The public witness is an extra, optional copy kept in an independent logbook.
Ship with proof, not hope
Free public beta for macOS and Linux. Point it at a repo and drive every agent you already use.