#1 Persistent memory for AI coding agents based on real-world benchmarks

Technical resources for AI developers to build applications, agents, and systems using Oracle AI Database and OCI services

Agent Memory Playground: AI Agent Memory Design & Optimization Techniques

QilbeeDB is a cloud-agnostic, high-performance graph database built for autonomous AI agents, real-time memory, and enterprise automation.

SmartAgentic is a project-centric, multi-agent memory OS for software engineering. It turns conversations and code decisions into an evolving knowledge graph so teams keep context across sessions, reduce contradictions, and ship.

Implementation based on the paper "ReasoningBank: Scaling Agent Self-Evolving with Reasoning Memory"

DevRouter — a collaborative, repository-aware context routing platform for AI coding agents that learns codebase structure and delivers the right context to the smallest capable model for faster, cheaper, and more accurate software development.

agentmemory: persistent memory for coding agents (Claude Code, OpenAI Codex, Cursor, Agent) with local markdown storage, daily logs, scratchpad, and qmd semantic search.

Aurora is a first-principles adaptive memory runtime organized around one evolving trace field.

A toy local memory system for multi-agent workflows

Pull-model episodic memory plugin for Hermes Agent. Real deletes, audit trace, BYO Claude. MIT.

Public fork of openai/codex with agentmemory-backed memory, expanded hook parity, and fork-focused runtime/docs cleanup.

A minimalist, cross-platform personal AI automation toolkit for research, learning, and lead generation. Built for Termux + WSL2 + Windows with Gemini CLI, AgentMemory, Apify, yt-dlp, and gcloud.

Public fork of rohitg00/agentmemory focused on Docker-first local deployment, Codex-native lifecycle ingestion, freshness-oriented retrieval, diagnostics.

AES-256-GCM encrypted memory for AI agents. Enterprise encryption layer for agentmemory.

🛠️ Experiment with AI agents in a local sandbox to run, score, and compare outputs without the need for complex frameworks.