Feature summary (what you would like to be able to do and where):
I would like the revert risk model that is currently used by Automoderator to also be available in RTRC (Real-Time Recent Changes). Specifically, it would be helpful if RTRC could display the revert risk score for each edit and allow patrollers to sort or filter based on that score.

Use case(s) (list the steps that you performed to discover that problem, and describe the actual underlying problem which you want to solve. Do not describe only a solution):
While patrolling recent changes in RTRC, I try to focus on edits that are most likely to be problematic. Currently, I rely on things like user status (TA vs registered), edit size, tags, and general experience.

At the same time, Automoderator already uses a revert risk model to assess edits automatically. That means we already have a machine learning signal that estimates how likely an edit is to be reverted but this information is not visible in RTRC.

The underlying issue is that automated tools and human patrolling tools are not using the same signals in a consistent way. The revert risk model exists and is actively used, but manual patrollers cannot benefit from it in RTRC (only in RecentChanges). This makes prioritization less efficient, especially on high-traffic projects.

Benefits (why should this be implemented?):

• Helps patrollers focus more quickly on high-risk edits.
• Makes better use of an already existing ML model.
• Improves consistency across moderation tools.
• Could reduce the time it takes for damaging edits to be reviewed or reverted.

(T408388)