#rag-memory · IA & ML
RAG & mémoire des IA
Chaînes de génération augmentée par recherche (RAG) et couches de mémoire qui permettent aux agents de se souvenir.
Ce mois-ci (4 oct. – 2 nov.), en tête : xerj (+16 204), Hy-Memory (+1 917) et survival-llm (+1 275). Les 433 projets classés ont gagné 264 359 étoiles.
Mis à jour le 2 nov., 12:33 UTC · classé toutes les heures à partir des étoiles quotidiennes de GitHub
Mis à jour il y a 1 h
Classement du mois
- 1xerj-org/xerj AI search engine and MCP server for bots and agents. One Rust binary indexes code, docs, logs, JSONL/CSV and PDFs for BM25 full-text, vector and hybrid search, code search without grep, agent memory and RAG. Elasticsearch-compatible API; Corpus Hub serves hundreds of reference memory corpora.Gagnées en 30 j +16,2 k Score 321,3 Total 18,7 k
- 2 Gagnées en 30 j +1,9 k Score 191,7 Total 1,9 k
- 3anonnerdfrenzy/survival-llm A free AI assistant and survival library that runs on your own computer, with no internet needed. Shows the source of every answer.Gagnées en 30 j +1,3 k Score 127,5 Total 1,3 k
- 4AgentMemoryRepo/agentmemoryrepo Spec for Agent Memory RepoGagnées en 30 j +844 Score 84,4 Total 844
- 5FengZhenfei/carrel Ontology-Augmented Generation for AI agents.Gagnées en 30 j +680 Score 68,0 Total 681
- 6HKUDS/DeepTutor DeepTutor: Lifelong Personalized Tutoring. https://deeptutor.info/.Gagnées en 30 j +13,5 k Score 66,9 Total 54,1 k
- 7akitaonrails/ai-memory Solution for long term memory for agent coding CLIs and to facilitate handoff between different agent vendorsGagnées en 30 j +6,1 k Score 65,6 Total 14,9 k
- 8 Gagnées en 30 j +637 Score 63,7 Total 637
- 9vectorize-io/hindsight Hindsight: Agent Memory That LearnsGagnées en 30 j +13,2 k Score 62,5 Total 57,9 k
- 10sahenjp/RustXNG Self-hosted, SSRF-hardened web search and grounding engine for local AI agents (AGPL-3.0-or-later)Gagnées en 30 j +615 Score 61,5 Total 615
- 11TencentCloud/Octop A smarter, self-hosted AI assistant — multi-user, multi-agent.Gagnées en 30 j +4,9 k Score 61,1 Total 11,5 k
- 12BotHarness/DeepSeekBot Open-source GrokBot alternative built on DeepSeek Harness. Build your personal bot team: multiple personal agents with persistent memory, identities and permissions, extensible through DSH plugins.|基于 DeepSeek Harness 的开源 GrokBot 平替:有长期记忆的个人 AI Bot 团队,支持多 Bot 协作、身份权限与插件生态扩展。MIT.Gagnées en 30 j +531 Score 53,1 Total 535
- 13Graphify-Labs/graphify Turn any codebase, with its docs, SQL schemas, configs, and PDFs, into a queryable knowledge graph. A /graphify skill for Claude Code, Cursor, Codex, and Gemini CLI: local deterministic AST parsing, every edge explained, no vector store.Gagnées en 30 j +17,9 k Score 51,2 Total 141 k
- 14tigerless-labs/agent-memory Long-term memory runtime for AI agents — plain Markdown as the source of truth, local ranked retrieval, and an independent sleep-time Manage layer. Claude Code and Codex share one store. No API key.Gagnées en 30 j +2,3 k Score 51,1 Total 4,3 k
- 15MemTensor/MemT-1.0 Official repository for MemT-1.0Gagnées en 30 j +498 Score 49,8 Total 498
- 16Graphicle-AI/graphicle Code graph and memory for your coding agents, in one Rust program. For Claude Code, Codex, Cursor and any MCP client.Gagnées en 30 j +493 Score 49,3 Total 493
- 17thedotmack/claude-mem Persistent Context Across Sessions for Every Agent – Captures everything your agent does during sessions, compresses it with AI, and injects relevant context back into future sessions. Works with Claude Code, OpenClaw, Codex, Gemini, Hermes, Copilot, OpenCode + MoreGagnées en 30 j +14,9 k Score 48,2 Total 110 k
- 18LazyAGI/LazyMind An AI Skill Runtime for knowledge-intensive work which makes everything under your control.Gagnées en 30 j +462 Score 46,2 Total 537
- 19FalkorDB/FalkorDB A super fast Graph Database uses GraphBLAS under the hood for its sparse adjacency matrix graph representation. Our goal is to provide the best Knowledge Graph for LLM (GraphRAG).Gagnées en 30 j +3,8 k Score 45,8 Total 10,5 k
- 20Yuuyuko-uu/dsh-fish-memory 🐋 DSH 鱼的记忆:给会话一份会淡忘的长期记忆,每条都带时间。名字里的「鱼」来自 DeepSeek 的鲸鱼。Gagnées en 30 j +455 Score 45,5 Total 456
- 21zgcagi/Harness-RSI Harness self-evolution: update targets, theory, evaluation, and reported evidence. Bilingual reading list and project website.Gagnées en 30 j +431 Score 43,1 Total 431
- 22wanghao-SunSky/cs329z-chinese 斯坦福 Stanford CS329Z Engineering AI Agents 中文教程:面向 Python 开发者的 AI 智能体课程,19 讲与 4 项实践,涵盖 LLM、RAG、工具调用、记忆、多智能体、评测与安全。完整 Markdown、在线与离线 HTML。Chinese course notes.Gagnées en 30 j +418 Score 41,8 Total 418
- 23DemonDamon/AgenticX AgenticX is a unified, production-ready multi-agent platform — Python SDK + CLI (agx) + Studio server + Machi desktop app. Features Meta-Agent orchestration, 15+ LLM providers, MCP Hub, hierarchical memory, avatar & group chat, skill ecosystem, safety sandbox, and IM gateway (Feishu/WeChat).Gagnées en 30 j +638 Score 40,9 Total 881
- 24zqiren/Orbital Context is yours. Agents are replaceable.Gagnées en 30 j +663 Score 37,8 Total 970
- 25Sallyn0225/eidolon-sc 和《偶像大师 闪耀色彩》偶像实时语音/文字对话的 AI 分身:日语原声 + 中文字幕、回忆原作剧情、跨角色共享记忆;支持 Google/GitHub 登录与自带密钥,Docker Compose 一键部署(参考 LAPLACE Eidolon)Gagnées en 30 j +367 Score 36,7 Total 367
- 26 Gagnées en 30 j +367 Score 36,7 Total 367
- 27sno-ai/sno-station When Claude Code hits its usage limit, Codex picks the task up from the same memory. Sno Station puts Codex, Claude Code, Cursor, Hermes Agent and OpenClaw on one squad: they hand work to each other and review each other's code. Open source, runs on your computer.Gagnées en 30 j +697 Score 35,1 Total 1,1 k
- 28aerovato/operator-memory The self-improving context engine for coding agents.Gagnées en 30 j +372 Score 34,0 Total 492
- 29aakarim/OpenLore A minimal, extensible, agent-native knowledge repository that keeps shared context current and inspectableGagnées en 30 j +888 Score 30,7 Total 1,7 k
- 30myneace/Gaussian_Clef_RAG Probabilistic retrieval-augmented generation on statistical manifolds with sub-100ms Cloudflare Clef decision routing, Woodbury-accelerated geometric kernels, and universal corpus generation. Every chunk of knowledge is a Gaussian distribution, not a flat vector. Retrieval is information geometry…Gagnées en 30 j +306 Score 30,6 Total 306
- 31Asymptote-Labs/agent-beacon The cross-harness, self-improving memory layer for AI agents. Join our community: https://discord.com/invite/zdNChS2fBuGagnées en 30 j +1,2 k Score 29,8 Total 3 k
- 321105216375-alt/resimind Neuro-symbolic agents with verified memory. MCP math tools, an algebra repair Skill, and an open challenge. Verify solutions, learn checked rules, and reuse.Gagnées en 30 j +294 Score 29,4 Total 294
- 33Sidiora-Labs/hyperfide The ai agent workflow tool that scales from small, simple projects to large, complex codebases.Gagnées en 30 j +426 Score 29,2 Total 639
- 34jaikydota/BullFlow 推量牛 BullFlow — 开源企业级 RAG 解决方案:从企业知识库、营销内容自动化生产到全域内容分发 | Open-source enterprise RAG solution: knowledge base → automated marketing content → omnichannel publishing. 知识库全链路本地开源模型,无需付费模型;AI Agent 创作图文 / 图片 / 视频,一键发布抖音、小红书、公众号、LinkedIn、YouTube 等 17 个平台。Commercial-grade, multi-tenant, billing & admin…Gagnées en 30 j +286 Score 28,6 Total 286
- 35666ghj/MiroFish A Simple and Universal Swarm Intelligence Engine, Predicting Anything. 简洁通用的群体智能引擎,预测万物Gagnées en 30 j +7,8 k Score 28,6 Total 83,4 k
- 36aoci-spec/aoci-code A persistent, Git-versioned map of your whole codebase and database schema that coding agents read before they touch anything. Local-first MCP server + CLI in Go: a governed repository index of code knowledge that gives Claude Code, Codex, Cursor, and opencode long-term context, memory, and code…Gagnées en 30 j +875 Score 28,2 Total 1,8 k
- 37ZipengWu365/TOMC TOMC Memory: CPU context compression and task memory for Codex, Claude Code, Cursor, TRAE, WorkBuddy and other MCP clients.Gagnées en 30 j +264 Score 26,4 Total 264
- 38ruvnet/ruflo 🌊 The original agent harness. Deploy intelligent multi-player swarms, coordinate autonomous workflows, and build conversational AI systems. Features adaptive memory, self-learning intelligence, federation, vector RAG integration, and native Claude Code / Codex / Hermes and many more IntegratedGagnées en 30 j +7 k Score 26,0 Total 80,6 k
- 39heaven999b/hello-agent-system 👋 Hello Agent System — learn enterprise AI agent system design by building every production layer from scratch. 17 bilingual lessons (EN/中文) with tested exercises: tools, architectures, reliability, security, evals, distributed execution, cost, RAG, release ops. 企业级 Agent 系统设计训练营Gagnées en 30 j +257 Score 25,7 Total 310
- 40zj-unicom-ai/UniEmployee 面向企业的数字员工构建与运行平台:把专业员工的经验、流程与判断标准,固化为可随时上岗、可配置、可审批、可观测的 AI 数字员工。Gagnées en 30 j +451 Score 25,6 Total 762
- 41SuanmoSuanyangTechnology/MemoryBear MemoryBear Equip AI with human-like memory capabilityGagnées en 30 j +2,2 k Score 25,4 Total 9,3 k
- 42metadist/synaplan-nextcloud Synaplan AI integration into nextcloud for RAG powered discussions, translations and summaries. Run your own AI platform next to your nextcloud!Gagnées en 30 j +253 Score 25,3 Total 255
- 43crow-hq/llm-wiki-v2 Turn your documents into a Markdown wiki you can read, that answers with citations. Works with any LLM provider, Ollama included. CROW classifier routing: 89% right note vs 62%, 2.1× faster ingest, 68% fewer tokens.Gagnées en 30 j +251 Score 25,1 Total 251
- 44Tencent/WeKnora Open-source LLM knowledge platform: turn raw documents into a queryable RAG, an autonomous reasoning agent, and a self-maintaining Wiki.Gagnées en 30 j +4,4 k Score 24,9 Total 36,2 k
- 45thomasunise/casecite CaseCite is an open-source, self-hosted legal RAG caselaw research & document analysis platform.Gagnées en 30 j +243 Score 24,3 Total 244
- 46metadist/synaplan Open-source, self-hosted AI platform (PHP/Symfony): chat with any model, RAG over your docs, embeddable widget, MCP — see the cost of every answer. Apache-2.0.Gagnées en 30 j +308 Score 23,0 Total 487
- 47SLEEPYBQ/MemoSync Co-manage memory with your coding agent across sessions. A local workbench for Claude Code, Codex and Pi.Gagnées en 30 j +222 Score 22,2 Total 222
- 48bojieli/ai-agent-book 《深入理解 AI Agent:设计原理与工程实践》(李博杰 著)开源主仓库:全书正文、编译版 PDF 与按章配套代码Gagnées en 30 j +4,9 k Score 21,4 Total 56,9 k
- 49JEen-wang/Adaptive-rag 按问题复杂度自动分流的电商客服 RAG:简单问题直接回答;复杂问题经过多跳检索并自我纠错;查单、退改通过工具调用完成,且高风险写操作必须先经人工确认。基于 FastAPI 和 LangGraph,混合检索并附带引用,内置离线评测与 Prometheus 指标。涵盖简历写法和项目面试拷打Gagnées en 30 j +207 Score 20,7 Total 207
- 50SourceWeft/SourceWeft Open-source, self-hostable alternative to ChatGPT and NotebookLM — agents, skills, MCP, and cited knowledge in one AI workstation. Your models, your infrastructure. Free to use GPT 6: https://SourceWeft.com/Gagnées en 30 j +238 Score 20,6 Total 372
1 dépôt à la croissance inhabituelle n’est pas affiché. Afficher le classement complet
Émergents en RAG & mémoire des IA Semaines passées : 26 oct.19 oct.