why - its awkward and inefficient to have skip logic spread across multi areas (custom policies rules and trusted client rules). We have to define the identifying criteria for safe clients in two places.
who - security engineers will benefit from this such as when they need to allow their dynamic scanner to bypass rules, like avoiding getting blocked by IP Reputation.
how - mimic what is done for WAF, MUM and BOT today. Just add another drop down that allows "skip" or "do not modify" for IP Reputation.