Oracle ships graph-aware memory framework for AI agents with typed relations
dr_cintas · x · 2026-10-15
Oracle Developers released Oracle AI Agent Memory, a database-backed Python memory framework that lets agents retrieve connected memories instead of flat text chunks.
Key mechanics:
- Links memory records with typed, directed relations: supersedes, refines, contradicts, supports, duplicates
- Builds a property graph inside the database so retrieval follows links rather than relying on vector similarity alone
- Trims retrieved results with a reranker and an LLM keep/discard pruner; expires old records via time-to-live settings
- Works with LangGraph, Claude Agent SDK, and OpenAI Agents SDK, with runnable notebooks on Oracle's GitHub
Install: pip install "oracleagentmemory==26.8.0". The payoff: agents get the latest fact and what changed it, not just the closest-sounding chunk.
Related event: Oracle Open-Sources Graph-Aware Memory Framework for AI Agents(3 posts)→
More from coding & agent
- Claude Code fast mode broken for 20x Max subscriber, user reports billing mismatch — RileyRalmuto · 2026-10-15
- Benchmarking coding agent harnesses and inference: lessons from $100Ks/month in research spend — abhijithneil · 2026-10-15
- AssetFurnace: Free Open-Source Local Image-to-3D Pipeline for Game Characters — chongdashu · 2026-10-15
- BaoCut's open-source approach to AI bilingual subtitle alignment, in five pitfalls — dotey · 2026-10-15
- Claude Code ships 278 changes in 2 days: 7 updates worth knowing — minchoi · 2026-10-15
- Video edited entirely by Claude Opus 4.5 High using free self-made skills, no editing software — chongdashu · 2026-10-15