Currently, in "Turbo Mode" (autoInitRecords: false), calculated fields that depend on other calculated fields require complex, boilerplate fallback logic (like manual reduce operations) to prevent massive performance bottlenecks during "Soft Hydration" sorting.

The Solution:

1. Engine Enhancement: Introduce a depends: [] config for Neo.data.Model fields.
2. Update Store.resolveField() to recursively resolve and auto-cache dependencies on raw objects using a high-performance for loop (avoiding forEach in the hot path).
3. Update the JSDoc in Model.mjs and Store.mjs to document this new architecture for the AI Knowledge Base.