[{"data":1,"prerenderedAt":4},["ShallowReactive",2],{"readme:memos":3},"\u003Cdiv align=\"center\">\n  \u003Ch1>\n    \u003Ca href=\"https:\u002F\u002Fmemos.openmem.net\u002F\" rel=\"nofollow ugc noopener\">\n      \u003Cimg src=\"https:\u002F\u002Fstatics.memtensor.com.cn\u002Flogo\u002Fmemos_color_m.png\" alt=\"MemOS Logo\" width=\"48\" \u002F>\n    \u003C\u002Fa> \n    MemOS 2.0 Stardust（星尘）\n  \u003C\u002Fh1>  \u003Cp align=\"center\">\n    \u003Ca href=\"https:\u002F\u002Fmemos-docs.openmem.net\u002Fhome\u002Foverview\u002F\" rel=\"nofollow ugc noopener\">\u003Cimg src=\"https:\u002F\u002Fimg.shields.io\u002Fbadge\u002FDocs-Get--Start-002FA7?labelColor=gray&style=for-the-badge&logo=googledocs&logoColor=white\" alt=\"Docs\" \u002F>\u003C\u002Fa>\n    \u003Ca href=\"https:\u002F\u002Farxiv.org\u002Fabs\u002F2507.03724\" rel=\"nofollow ugc noopener\">\u003Cimg src=\"https:\u002F\u002Fimg.shields.io\u002Fbadge\u002FArXiv-2507.03724-B31B1B?labelColor=gray&style=for-the-badge&logo=arxiv&logoColor=white\" alt=\"ArXiv\" \u002F>\u003C\u002Fa>\n    \u003Ca href=\"https:\u002F\u002Fx.com\u002FMemOS_dev\" rel=\"nofollow ugc noopener\">\u003Cimg src=\"https:\u002F\u002Fimg.shields.io\u002Fbadge\u002FFollow-MemOS-000000?labelColor=gray&style=for-the-badge&logo=x&logoColor=white\" alt=\"X\" \u002F>\u003C\u002Fa>\n    \u003Ca href=\"https:\u002F\u002Fdiscord.gg\u002FTxbx3gebZR\" rel=\"nofollow ugc noopener\">\u003Cimg src=\"https:\u002F\u002Fimg.shields.io\u002Fbadge\u002Fdynamic\u002Fjson?url=https%3A%2F%2Fdiscord.com%2Fapi%2Fv10%2Finvites%2FTxbx3gebZR%3Fwith_counts%3Dtrue&query=%24.approximate_presence_count&suffix=%20online&label=Discord&color=404EED&labelColor=gray&style=for-the-badge&logo=discord&logoColor=white\" alt=\"Discord\" \u002F>\u003C\u002Fa>\n    \u003Cbr \u002F>\n    \u003Ca href=\"https:\u002F\u002Fgithub.com\u002FIAAR-Shanghai\u002FAwesome-AI-Memory\" rel=\"nofollow ugc noopener\">\u003Cimg src=\"https:\u002F\u002Fimg.shields.io\u002Fbadge\u002FResources-Awesome--AI--Memory-8A2BE2?labelColor=gray&style=for-the-badge&logo=awesomelists&logoColor=white\" alt=\"Resources\" \u002F>\u003C\u002Fa>\n  \u003C\u002Fp>  \u003Cp align=\"center\">\n    \u003Cstrong>Give your Agent persistent memory and the ability to grow.\u003C\u002Fstrong>\u003Cbr \u002F>\n  \u003C\u002Fp>  \u003Cp align=\"center\">\n    \u003Cstrong>English\u003C\u002Fstrong> | \u003Ca href=\"https:\u002F\u002Fgithub.com\u002FMemTensor\u002Fmemos\u002Fblob\u002FHEAD\u002FREADME_ZH.md\" rel=\"nofollow ugc noopener\">中文\u003C\u002Fa>\n  \u003C\u002Fp>\n\u003C\u002Fdiv>\u003Cdiv align=\"center\">\n  \u003Cimg width=\"1660\" alt=\"MemOS agent ecosystem: OpenClaw, Hermes, and DeepSeek Harness\" src=\"https:\u002F\u002Fraw.githubusercontent.com\u002FMemTensor\u002Fmemos\u002FHEAD\u002Fassets\u002Freadme\u002Fmemos-agent-ecosystem.png\" \u002F>\n\u003C\u002Fdiv>\u003Cblockquote>\n\u003Cp>[!TIP]\n\u003Cstrong>New: Connect MemOS to DeepSeek Harness (\u003Ccode>dsh\u003C\u002Fcode>)\u003C\u002Fstrong>\u003C\u002Fp>\n\u003Cp>Add automatic recall, background capture, hybrid retrieval, and a local Memory Viewer to DeepSeek Harness—powered by the same MemOS core used across agent ecosystems.\u003C\u002Fp>\n\u003Cp>\u003Cstrong>\u003Ca href=\"#memos-plugin\" rel=\"nofollow ugc noopener\">Get started →\u003C\u002Fa>\u003C\u002Fstrong>\u003C\u002Fp>\n\u003C\u002Fblockquote>\n\u003Chr \u002F>\n\u003Ch2>👾 MemOS: Memory Operating System for LLM & AI Agents\u003C\u002Fh2>\n\u003Cp>\u003Cstrong>MemOS\u003C\u002Fstrong> is a Memory Operating System for LLMs and AI agents that unifies \u003Cstrong>store \u002F retrieve \u002F manage\u003C\u002Fstrong> for long-term memory, enabling \u003Cstrong>context-aware and personalized\u003C\u002Fstrong> interactions with \u003Cstrong>KB\u003C\u002Fstrong>, \u003Cstrong>multi-modal\u003C\u002Fstrong>, \u003Cstrong>tool memory\u003C\u002Fstrong>, and \u003Cstrong>enterprise-grade\u003C\u002Fstrong> optimizations built in.\u003C\u002Fp>\n\u003Ch3>Key Features\u003C\u002Fh3>\n\u003Cul>\n\u003Cli>\u003Cstrong>Unified Memory API\u003C\u002Fstrong>: A single API to add, retrieve, edit, and delete memory—structured as a graph, inspectable and editable by design, not a black-box embedding store.\u003C\u002Fli>\n\u003Cli>\u003Cstrong>Multi-Modal Memory\u003C\u002Fstrong>: Natively supports text, images, tool traces, and personas, retrieved and reasoned together in one memory system.\u003C\u002Fli>\n\u003Cli>\u003Cstrong>Multi-Cube Knowledge Base Management\u003C\u002Fstrong>: Manage multiple knowledge bases as composable memory cubes, enabling isolation, controlled sharing, and dynamic composition across users, projects, and agents.\u003C\u002Fli>\n\u003Cli>\u003Cstrong>Asynchronous Ingestion via MemScheduler\u003C\u002Fstrong>: Run memory operations asynchronously with millisecond-level latency for production stability under high concurrency.\u003C\u002Fli>\n\u003Cli>\u003Cstrong>Memory Feedback & Correction\u003C\u002Fstrong>: Refine memory with natural-language feedback—correcting, supplementing, or replacing existing memories over time.\u003C\u002Fli>\n\u003C\u002Ful>\n\u003Ch3>News\u003C\u002Fh3>\n\u003Cul>\n\u003Cli>\u003Cp>\u003Cstrong>2026-08-17\u003C\u002Fstrong> · 🐋 \u003Cstrong>MemOS Connects with DeepSeek Harness\u003C\u002Fstrong>\nMemOS now brings persistent memory to \u003Cstrong>DeepSeek Harness\u003C\u002Fstrong> through both local and cloud plugins. DSH can automatically recall relevant context before a task and retain new experience after a successful turn, without modifying its core.\u003C\u002Fp>\n\u003C\u002Fli>\n\u003Cli>\u003Cp>\u003Cstrong>2026-07-02\u003C\u002Fstrong> · 🏆 \u003Cstrong>MemOS Advances Agent and User Memory Benchmarks\u003C\u002Fstrong>\nWith MemOS, \u003Cstrong>OpenClaw\u003C\u002Fstrong> improves average task completion from \u003Cstrong>36.63% to 50.87%\u003C\u002Fstrong> across five agent tasks. MemOS also achieves \u003Cstrong>88.83 on LoCoMo\u003C\u002Fstrong> and \u003Cstrong>89.20 on LongMemEval\u003C\u002Fstrong>, and leads in \u003Cstrong>OmniMemEval\u003C\u002Fstrong>, a unified evaluation of 14 commercial memory products across ten datasets.\u003C\u002Fp>\n\u003C\u002Fli>\n\u003Cli>\u003Cp>\u003Cstrong>2026-05-09\u003C\u002Fstrong> · 🧠 \u003Cstrong>memos-local-plugin 2.0\u003C\u002Fstrong>\u003C\u002Fp>\n\u003C\u002Fli>\n\u003C\u002Ful>\n",1790587591161]