Proposal: memory observability and evaluation for agent workflows

Hi, I’m Vivek Gupta, Founder & CEO at [MemCode](https://memcode.in). AgentOps gives teams visibility into agent behavior, and durable memory introduces important signals beyond traces: what was written, what was retrieved, whether it was relevant, and whether it improved the run.

Would AgentOps consider an optional MemCode integration? MemCode could provide the memory layer while AgentOps captures memory read/write events, scope, latency, retrieval relevance, and downstream outcomes for debugging and evaluation.

I’d be glad to prepare a small instrumentation example or PR around the event hooks you recommend. The integration can remain optional and should preserve existing observability workflows. Is there a preferred technical discussion channel for proposing memory-specific events?