Excessive retry backoff (~2 weeks) on "Quota Exceeded" with no mechanism to reset/force retry #13591

Question

I encountered a Too Many Requests: quota exceeded error while using the claude-opus-4.6 model due to insufficient funds. The tool triggered a retry mechanism with an unreasonably long wait time (~2 weeks).

After replenishing my GitHub funds/credits, the tool remains stuck in this retry loop. I cannot find a way to manually force a retry or reset the timer without potentially losing my current context/session.

Steps to Reproduce:

Run a task using claude-opus-4.6.

Exhaust the API quota/funds to trigger a 429 Too Many Requests error.

Observe the retry timer.

Current Behavior:

The CLI displays:

Too Many Requests: quota exceeded [retrying in ~2 weeks attempt #1]
The application waits indefinitely (or for the displayed duration) and does not detect that funds have been restored.

Expected Behavior:

The retry backoff time should be reasonable (e.g., minutes, not weeks).

There should be a keybinding or command to force an immediate retry (e.g., r to retry) after the user has resolved the payment issue.

Screenshots:

Environment:

Model: Claude Opus 4.6 (via GitHub Copilot)

OS: Linux