MEMORY

A late-July research wave - native in-backbone memory, pretrained parametric memory at scale, memory reconstruction, and transactional memory writes - challenges the external-store paradigm every agent memory product is built on. Here is what changes by late 2027 and what developers should do now.

A fair, sourced comparison of the memory layers developers reach for in 2026: Mem0's extract-and-retrieve, Zep's temporal knowledge graph, Letta's self-editing agent memory, and Cloudflare's Durable Objects primitive. Architecture, pricing, the benchmark disputes, and which to pick for your agent.

Configurable memory, sandbox-aware orchestration, Codex-like filesystem tools. Here is how the new Agents SDK actually behaves in prod.

Cloudflare's Agent Memory primitive. What it stores, latency profile, how it compares to mem0, and how to wire it into your stack.

Define custom subagent types within your project's memory layer.

Automatic session-to-session memory of build commands, errors, and learnings.

View and edit auto-memory and CLAUDE.md via the /memory command.

Configure model, tools, MCP, skills, memory, and scoping.

Auto-memory that persists across multiple subagent invocations.

AI agents that reflect on failures, accumulate skills, and get better with every session. Reflection patterns, memory architectures, skill extraction, and working code examples for building agents that actually learn.

AI-powered context manager that remembers your code, snippets, links, and project context across IDEs, browsers, and terminals. Local-first with 9 months of memory. Free.

Agents forget everything between sessions. Here are the patterns that fix that: CLAUDE.md persistence, RAG retrieval, context compression, and conversation summarization.

Skills turn Claude Code sessions into persistent memory. Successes and failures get captured, progressively disclosed, and shared across teams. Your agent remembers.