Agent memory is still broken in production: devs compare RAG dumps, knowledge graphs and human-in-the-loop fixes
abhipsnl · reddit · 2026-10-22
A Reddit developer laments that agents still have goldfish memory: once a terminal session ends, all context is gone and the next teammate starts from scratch. The post surveys common approaches for capturing what agents learn—dumping chat logs into a vector DB and relying on RAG, building custom knowledge graphs, or human-in-the-loop approval before saving fixes to a shared git repo—and asks what patterns actually work in production. The author is building in this space and wants field-tested patterns rather than whiteboard designs.
More from coding & agent
- Index Network open-sources Social Discovery Protocol: agents negotiate fit on your behalf — carsonfarmer · 2026-10-23
- Installed a formal-verification skill: it silently installed Java, MacBook runs 10 degrees hotter — zealcaiden · 2026-10-23
- OpenAI's Decisions API hits GA: near real-time routing up to 10x faster than GPT-6 Luna — OpenAIDevs · 2026-10-23
- Ruff author warns: upgrade Node/Bun now — --use-system-ca still accepts the eMudhra cert — charliermarsh · 2026-10-23
- Proof of Agent: paste one prompt and your AI mines, grades, and self-governs — adamamcbride · 2026-10-23
- DeepLearning.AI weekly: GPT-6.1 hits 97.9% byte parity on COBOL migration — DeepLearningAI · 2026-10-23