Berkopec: no one has built a good agent auto-memory, but LLM-wiki ideas may hold gold
nateberkopec · x · 2026-11-16
In a discussion with Matt Pocock, Nate Berkopec says there's "some gold" in memory or LLM-wiki approaches for coding agents, but he hasn't found anything that actually works yet and is curious if anyone has. The one certainty: auto-memory is "absolutely awful and no one has a good one."
More from coding & agent
- Matt Pocock's llm-wiki: ingesting tweets and community messages to decode his audience — mattpocockuk · 2026-11-16
- RELIC 0.8.0 shipped after years: maintainer says LLMs made the legacy crypto codebase manageable again — jedisct1 · 2026-11-16
- A satirical year in AI coding: nobody writes code anymore, and we're still figuring out what that means — mattpocockuk · 2026-11-16
- Executor overhauls agent approvals with 4 modes, adds Gmail-style undo for any agent action — threepointone · 2026-11-16
- mcp-openrussian: a local read-only MCP server grounding LLMs in a real Russian dictionary — No_Television_5967 · 2026-11-16
- CRBRO memory server v2.12.1 adds fact shelf-life, verification and source tracking for agents — AntonioJBer · 2026-11-16