memory-cloud

<p align="center"> </p>

<p align="center"> <strong>Universal AI Memory Platform</strong> — Self-hosted, open source.<br> Give your AI assistants persistent memory across conversations. </p>

<p align="center"> <a href="LICENSE"></a> <a href="https://github.com/kagura-ai/memory-cloud/actions/workflows/ci.yml"></a> <a href="https://codecov.io/gh/kagura-ai/memory-cloud"></a> <a href="https://www.python.org/downloads/"></a> <a href="https://nodejs.org/"></a> <a href="https://modelcontextprotocol.io/"></a> </p>

<p align="center"> Works with Claude, ChatGPT, Gemini, and any MCP-compatible client.<br> <a href="https://github.com/kagura-ai/kagura-memory-python-sdk"><strong>Python SDK (KaguraClient / KaguraAgent)</strong></a> </p>

Kagura Memory Cloud is a memory system that lets AI assistants remember, search, and learn from past conversations. It provides:

• 10 MCP Tools — remember, recall, forget, reference, explore, get_context_info, list_contexts, create_context, update_context, usage_guide
• 3-Layer Memory Architecture — summary (search-optimized) / context / full content
• Hybrid Search — 60% semantic (OpenAI embedding) + 40% full-text (Qdrant BM25)
• Neural Memory — Hebbian learning creates automatic connections between related memories
• Team Collaboration — Workspaces with RBAC (Owner/Admin/Member/Viewer)
• **W