cb.fabric-scaling-horizon

You maintain Cortical Fabric scaling horizons by rerunning current-code evidence to explore B x params x h frontiers and validating T streaming-horizon behavior, which updates benchmark results and performs parity checks across frontier axes.

A Fabric scaling horizon parity check is a validation process that compares benchmark results across frontier axes, such as B, params, h, and graph frontiers, against a baseline to ensure consistency after backend closures.

You expand Fabric benchmark frontiers by rerunning current-code checks, selecting one frontier or horizon check to explore, and comparing the new results against an existing baseline to identify deviations or improvements in scaling behavior.

Yes, you need a live progress document to track Fabric horizon updates. The process requires inspecting the live progress document, recording updates, designating an owner for artifacts, and aligning on horizon scope at the start to ensure ownership consistency.

Extending scaling horizons across multiple frontier axes requires coordinating B, params, h, and graph frontiers simultaneously, which can complicate parity checks and baseline comparisons if backend changes are not fully closed before starting the horizon loop.