Top MEMORY GitHub Repositories & Tools (2026)

Oracle Coherence Community Edition

Open-source, Git-native memory for AI coding agents — deterministic local recall for Claude Code, Codex, Cursor, Gemini CLI, and compatible tools.

Config-driven engine that turns JSON into production-grade AI agents. Multi-agent orchestration, 12+ LLM providers, MCP/A2A protocols, RAG, persistent memory, and enterprise compliance (EU AI Act, GDPR, HIPAA). Built on Quarkus.

High-Performance Java-Native-Persistence. Store and load any Java Object Graph or Subgraphs partially, Relieved of Heavy-weight JPA. Microsecond Response Time. Ultra-High Throughput. Minimum of Latencies. Create Ultra-Fast In-Memory Database Applications & Microservices.

AI-Native & Cloud-Native FS: A high-performance file semantic layer for cloud object storage, integrated with high-speed cache. CNCF Sandbox Project.

Capture of the local Among Us executable state

Distributed, transactional document database backed by FoundationDB.

Opensource Memory for Agents

High Performance KV Cache for LLM Inference

Long-term memory for Hermes and Codex, with local storage, source-backed recall, and tools to inspect, correct, or delete remembered information.

A secure, high-performance, multi-tenant platform for remote App ”Hosting“ and ”Computing“

A rclcpp-compatible true zero-copy IPC middleware that supports all ROS message types, including message structs already generated by rosidl.

Tinybot is a lightweight personal AI Agent that is constantly evolving

Open-source performance diagnostics for PyTorch training runs.

OWASP Foundation web repository

Local-first, inspectable memory for coding agents: durable context across sessions and repositories, code-aware recall, bi-temporal history, MCP, and a self-hosted WebUI.

Zerocopy makes zero-cost memory manipulation effortless. We write `unsafe` so you don’t have to.

Low-latency java socket implementation (using shared memory)

SpeakerMem-R1: Speaker-Centered Dual-Track Memory for Multi-Party Dialogue

Full AI context and content layer for coding agents over one MCP server — tree-sitter code-map, document RAG, shared memory, multi-agent comms, web crawl, git history + blame. 300+ languages, 10+ agent harnesses, pure Rust.