cavemem: Cross-Agent Persistent Memory for Coding Assistants, Stored Compressed and Local-First

GitHub·low signal

From the same author as the viral caveman skill, cavemem is a local-first, compressed persistent-memory layer meant to be shared across coding assistants rather than locked to one agent. It targets the real gap of memory that doesn't survive across tools and sessions. Early and single-source on adoption, so low confidence on traction, but the cross-agent, local-by-default design is a sensible response to portable-memory demand.