A deep dive into LLM agent memory architecture based on the MemRec paper (ACL 2026): collaborative memory, decoupling memory management from reasoning, Information Bottleneck. Comparison of the TOP-3 production tools — Mem0, Letta (MemGPT), and Zep — across nine axes. Highlighting the gap: research outpaces tooling.
Deep dive into agent memory mechanics: five memory files, four context assembly strategies, TOOLS.md as a dual-source responsibility, and Dreaming — background consolidation from Anthropic (Auto Dream, Dreams API, KAIROS) and OpenAI (Dreaming V3, 82.8% factual recall). Engineering decisions behind a production platform.
Five multi-agent orchestration architectures — Supervisor, Swarm, Debate, Assembly Line, Hierarchical — with Mermaid diagrams, code examples in Eino (Go), and an honest look at limitations. When one agent isn’t enough, and when you’re better off staying solo.
Context window ≠ memory. I break down agent amnesia, the academic taxonomy (Du et al., 2026), five approaches to memory — from OpenClaw’s file-based brain to mem0’s managed layer — and how our team solves this in practice.
Plan-and-Execute separates strategic planning from tactical execution. I break down the Planner+Executor+Reviser architecture, ReWOO and LLMCompiler variants with numbers from original papers, prompt injection defense via control-flow integrity, and a minimal working Go example with Eino.