Collected sources and patterns will appear here. Add from search, explore, or the patterns library.
CacheObject -> PinnedCacheObject
Protect critical cached segments from eviction during hierarchical memory swaps using explicit soft-pin or hard-pin lifecycle states.
Problem it solves
Generic LRU eviction can discard warm reusable KVCaches or model weights, causing redundant generation/re-computation.
Consumes
Emits
The real projects this mechanism was found in. Attribution is the point — this is how the best teams actually do it.