fix: retry failed HTTP calls and reduce check batch size

[reheatvet]

What's wrong

On my host, the SDK's HttpClient drops requests when too many are fired at once. One failed call crashes the whole cycle because
Promise.all rejects on any single failure. The node gets stuck in a crash-retry loop every 5 minutes.

What this does

• Adds retry (3 attempts with backoff) to the base function so transient network failures don't kill the cycle. Reverts aren't retried.
• Drops CHECK_BATCH from 50 to 10 so fewer calls are made at once.

Results

Tested on my host that previously crashed every cycle:

• Before: crash loop, never completes
• Retry only: works but takes ~4.5 min
• Retry + smaller batches: ~28s per phase, no issues

Hope this is fine, let me know if you'd like any changes