Local persistent memory store for LLM applications including claude desktop, github copilot, codex, antigravity, etc.

For Claude.

Neura lets you run long, tool heavy coding sessions without blowing up token costs by compressing old context into references and only restoring exact data when needed, reducing hallucinations by grounding the model in real, retrievable source data instead of guesswork.

Semantic retrieval for AI agents: high-recall memory and code search that cuts coding-agent token usage by nearly 50%

MCP Context Server — a FastMCP-based server providing persistent multimodal context storage for LLM agents.

Durable hybrid memory for AI agents: vector + BM25 + temporal + graph recall, exposed through MCP, HTTP, and Python.

把单个 AI 助手升级为可扩展的多专家协作运行时， 带工业级记忆检索、会话管理和完整的开发者可观测性。

Provide precise OpenClaw documentation and exact configs to your coding agent for accurate answers without guesswork or summaries.

Memory Utility Networks (MUN): A research-oriented framework for utility-based memory retrieval and example selection in long-horizon decision making and in-context learning. Includes reproducible experimentation, validation audits, and ablation studies investigating learned memory utility.

Minimal starter for hybrid memory retrieval in OpenClaw-style agent workflows.

mRAG (micro-RAG) is an agnostic backend memory encoder and retrieval system designed to segment memories into basic short form belief statements that can be retrieved and injected directly into an Agent's context using multi-head queries to inject only the most highly relevant beliefs with minimal excess.

Replication of Recanatesi et al. (2015) https://www.frontiersin.org/articles/10.3389/fncom.2015.00149/full

Enable persistent memory for Claude that works across crashes and restarts in both Claude Code and Claude Cowork environments.

📝 Create and manage your notes with Memos, the open-source, self-hosted service that ensures your data stays private and in your control.

Scripts used to analyze behavioral responses during memory retrieval and the association between fMRI-BOLD activation during memory encoding and behavioral responses during memory retrieval described in Ruiz-Rizzo et al. (2022), Archives of Cinical Neuropsychology

Local-first AI memory store with synaptic dynamics (strengthen/decay/prune) and neighbor-based pattern expansion (L2).

ChronoMind is a temporal memory retrieval system that combines Qdrant vector search, Neo4j graph traversal, BM25 lexical retrieval, learned ranking, and timeline reconstruction to search personal memories with causal and temporal reasoning.