#memory

When I started learning Redis, I had one big question: "If thousands of people are using the same app...

I've been building agents for a while and one thing keeps bugging me: for a tool-using agent (ReAct /...

The Slowest Incident to Diagnose Memory leaks are sneaky. The service runs fine for hours....

Yesterday I closed a paper tab and sat with a realization for a moment. I'd just finished reading...

Deterministic consolidation, a typed graph in SQLite, and an honest look at what agent-memory...

Yes — several MCP servers exist for agent memory: the official knowledge-graph reference server, PLUR, Zep, Mem0's OpenMemory, and community projects. MCP standardizes the transport; the memory format is the differentiator.

VetoBench measures the one thing memory benchmarks skip: when a task invites an approach your team already ruled out, does the agent propose it again? Here's how RoBrain and Mem0 did on identical input.

The setup 33 autonomous bots, 43 timers, persistent state via ChromaDB + JSONL + SQLite...

Fine-tuning bakes facts into weights you can't inspect, delete, or port — and every update pays the parallel learning tax. For teaching an AI facts that change, an external memory layer wins on cost, auditability, and portability.

Anyone who has used AI tools for a while has probably run into this annoyance. You ask it to write a...

PVM: Give any LLM long-term memory. No API keys, no GPU, ~800 lines Python. Every LLM...

Agent memory isn't a retrieval problem. It's a forgetting problem, and almost nobody is building the part that decides what to keep.