What is the reason this setting causes rate limiting on the /v2/wfm/schedules endpoint when querying agent availability for bot fallback logic? The integration hits 429s consistently during UK business hours despite caching layers, suggesting the cache invalidation trigger is misaligned with the WFM sync interval or the API gateway is throttling based on source IP rather than token. Need to know if there is a specific header or payload optimization required to bypass this saturation.