I think the CoALA paper's classification system of semantic/episodic/procedural is maybe the closest thing we have to a standard for agent m…

The CoALA paper proposes a classification system for agent memory that distinguishes between semantic memory (facts and concepts), episodic memory (specific experiences and events), and procedural memory (skills and how-to knowledge), which Harrison Chase suggests represents a meaningful standardization framework for conceptualizing different types of agent memory. This classification system provides a structured approach for understanding how AI agents organize and utilize different categories of information.

very memgpt / sarah wooders coded. memory isn’t a layer, it is the system. most teams think they’re choosing a model, but they’re really cho…
'every memory system is choosing a position on the raw/derived spectrum. and neither extreme works.' the axis nobody talks about: ownership …
your harness = your memory if you'd rather read in blog form, link here: https://blog.langchain.com/your-harness-your-memory/
This is the right frame. We’re currently designing our agent memory platform, and the hardest part isn’t storage — it’s deciding what to rem…
Agent harnesses dominate agent building and tie intimately to memory. Closed harnesses behind proprietary APIs force yielding control of age…
The ‘agent harness’ is essentially the new runtime. It’s like a lightweight OS around an LLM—handling memory, tool calls, retries, and contr…

Source: Harrison Chase (X) | 2026-04-21

Related