Cohort of 83 admitted agents tagged capability:memory. Composite below is the cohort's average AgentScore.
| Cmp | Rank | Agent | 24h | Score | Δ24h | Watch |
|---|---|---|---|---|---|---|
| #11 | Understand-Anything mitcliUnderstand-Anything: Graphs that teach > graphs that impress. Turn any code, or knowledge base (Karpathy LLM wiki), into an interactive knowledge graph you can explore, search, and ask questions about. Works with Claude Code, Codex, Cursor, Copilot, Gemini CLI, and more. | 60.3 | +0.00 | |||
| #46 | letta apache-2.0letta: Letta is the platform for building stateful agents: AI with advanced memory that can learn and self-improve over time. | 7 | 51.4 | +1.67 | ||
| #50 | GenericAgent mitGenericAgent: Self-evolving agent: grows skill tree from 3.3K-line seed, achieving full system control with 6x less token consumption | 50.5 | +0.01 | |||
| #69 | ai-memory mitcliai-memory: Solution for long term memory for agent coding CLIs and to facilitate handoff between different agent vendors. | 46.8 | -0.08 | |||
| #85 | evolver gplmcp-serverevolver: The GEP-powered self-evolving engine for AI agents. Auditable evolution with Genes, Capsules, and Events. | evomap.ai | 1 | 42.6 | +0.02 | ||
| #92 | zerostack gplzerostack: Minimalistic coding agent written in Rust, optimized for memory footprint and performance. | 40.6 | +0.02 | |||
| #110 | opensquilla apache-2.0mcp-serveropensquilla: OpenSquilla — Token-Efficient AI Agent with same budget, higher intelligence density. | 1 | 37.1 | +0.02 | ||
| #125 | memory-os mitide-pluginmemory-os: A 6-layer memory operating system for Hermes Agent — persistent memory with Qdrant, structured facts, fabric recall, auto-curated wiki, and surgical context injection. Runs locally, any LLM provider. | 1 | 35.7 | -0.01 | ||
| #130 | PaperGuru-Benchmark PaperGuru-Benchmark: Lifecycle-Aware Memory for long-horizon LLM agents — 66.05% on PaperBench, 94.66% on SurveyBench, 10 peer-reviewed acceptances at FSE/ICML/TOSEM/AEI/ICoGB | 1 | 35.3 | +0.05 | ||
| #136 | munder-difflin munder-difflin: local multi-agent harness. | 1 | 34.8 | -0.01 | ||
| #225 | elephant-agent elephant-agent: Personal-Model First Self Evolving AI Agent 🐘. | 4 | 27.9 | -0.01 | ||
| #254 | ai-agents-from-scratch mitai-agents-from-scratch: Demystify AI agents by building them yourself. Local LLMs, no black boxes, real understanding of function calling, memory, and ReAct patterns. | 5 | 26.6 | +0.00 | ||
| #269 | Photo-agents mitPhoto-agents: Autonomous self-evolving agents. Vision-grounded layered memory and self-written skills for LLM agents that operate your computer. | 1 | 25.6 | +0.00 | ||
| #286 | mempalace-evolve mitmempalace-evolve: Self-evolving memory palace for AI agents — persistent memory with automatic learning, knowledge graph, and multi-agent support. | 42 | 25.1 | -1.74 | ||
| #293 | HtooTayZa/sawtooth-memory HtooTayZa/sawtooth-memory: discovered AI agent. | 24.8 | -0.00 | |||
| #294 | TriMem mitTriMem: [arXiv:2605.19952] "Rethinking How to Remember: Beyond Atomic Facts in Lifelong LLM Agent Memory" | 24.8 | 0.00 | |||
| #298 | juice mcp-serverjuice: A negative-constraint memory layer for AI agents. | 1 | 24.8 | 0.00 | ||
| #305 | Drone-Rental-System apache-2.0mcp-serverDrone-Rental-System: 该项目是无人机的租赁系统，也可以是其他的设备、器材租赁，技术栈：Java 21 + Spring Boot 3.5.13 + Vue 3 + Spring AI + MCP + AI Memory，其他的技术栈Dify/Coze、 Python LangChain/LangGraph Agent 。项目是开源的，免费学习的，star支持一下，不断更新中。 | 2 | 24.4 | -0.08 | ||
| #312 | riskkernel apache-2.0mcp-serverriskkernel: Deterministic cost / loop / time budgets · full observability · crash-resumable runs · human-approval gates · a memory you own. Self-hosted. Your keys. No telemetry. Point it at your existing agents - one env var. | 1 | 24.1 | 0.00 | ||
| #329 | Infini-Memory apache-2.0Infini-Memory: Infini-Memory: A maintainable, text-based persistent memory architecture that organizes LLM agent memory as topic-structured documents. paper: https://arxiv.org/abs/2606.10677 | 23.4 | +0.02 | |||
| #332 | sawtooth-memory mitsawtooth-memory: Async hierarchical memory middleware for LLM agents. | 2 | 23.2 | -0.00 | ||
| #336 | compass-skills mitcompass-skills: 司南：个性化 AI 任务总控 Skills 系统 /COMPASS: Personal Alignment Skills OS for AI Agents. | 4 | 23.0 | -0.12 | ||
| #338 | ruvos mcp-serverruvos: rUvOS — the agentic operating system. A single static Rust binary that runs as an MCP server, giving Claude Code persistent memory, resumable signed sessions, multi-agent orchestration, a knowledge graph, safety guardrails, and a signed audit trail. Zero Node.js, zero SQLite,... | 3 | 23.0 | -0.07 | ||
| #352 | llm-wiki_obsidian_hermes_r0b0tlabbra1n mitllm-wiki_obsidian_hermes_r0b0tlabbra1n: Filesystem-first LLM-Wiki + Obsidian + Hermes Agent memory system. Markdown source of truth, SQLite FTS5 search, secret scanning, tier-based memory. Built for local LLM setups. | 2 | 22.5 | 0.00 | ||
| #364 | agent-harness-generator mitmcp-serveragent-harness-generator: 🛠️ The meta-harness for AI agents — scaffold your own focused, branded agent harness with its own npx CLI, MCP server, memory, learning loop, and witness-signed releases. Works with Claude Code, Codex, pi.dev, Hermes, OpenClaw, and RVM (hardware-isolated). | 3 | 22.2 | +0.09 | ||
| #366 | vibecode-pro-max-kit mitclivibecode-pro-max-kit: RIPER-5, spec-driven, context memory self-improving agent harness for Claude Code & Codex — 12 specialist agents, 31 skills, spec-driven development with phase-locked safety. Works with any codebase. Install in 30 seconds. Community skills & agents welcome. Built for vibecoder... | 2 | 22.2 | -0.02 | ||
| #373 | Agent_Memory_Techniques apache-2.0Agent_Memory_Techniques: Agent memory for LLMs: 30 runnable Jupyter notebooks covering conversation buffers, vector stores, knowledge graphs, episodic and semantic memory, MemGPT, Mem0, Letta, Zep, Graphiti, LoCoMo benchmarks, and production patterns. | 5 | 21.8 | +0.04 | ||
| #391 | memwatch mitmemwatch: Memory staleness detection and auto-expiry for LLM agents. | 6 | 21.4 | +0.11 | ||
| #416 | agent-oss apache-2.0ide-pluginagent-oss: A recursive evidence-gated cognitive runtime for memory-native AI agents, combining hybrid retrieval, temporal reasoning, async learning, and plug-and-play tools. | 2 | 20.8 | +0.01 | ||
| #426 | Awesome-Personalization-in-MLLMs Awesome-Personalization-in-MLLMs: Awesome papers on personalization in LLMs/MLLMs: memory, alignment, and evaluation. | 1 | 20.5 | 0.00 | ||
| #456 | personal-agent-template mitpersonal-agent-template: Open-source template for a durable personal AI agent — web chat, Slack, Linear, and long-term memory with user-approved saves. Eve, Nuxt, Better Auth, Vercel Connect. | 101 | 19.7 | -2.73 | ||
| #458 | kaeru mitkaeru: Cognitive memory layer for LLM agents. | 23 | 19.7 | +0.53 | ||
| #475 | bazi-ziwei-skill mitbazi-ziwei-skill: AI 八字 + 紫微斗数排盘与综合印证 Skill：算法精准排盘（不靠 LLM 猜），三种分析模式，一键生成水墨风 HTML 命盘海报。兼容 Claude / Codex / Cursor / Workbuddy 等 SKILL.md Agent。 | 114 | 19.4 | -2.96 | ||
| #502 | Analogagent mitlibraryAnalogagent: Training-free agentic framework for analog circuit design automation: an LLM multi-agent system with self-evolving memory. KDD 2026 (AI4Science, Oral). | 2 | 18.8 | 0.00 | ||
| #513 | piia-engram apache-2.0mcp-serverpiia-engram: Your AI identity layer — one identity, every tool. Local-first, MCP-compatible, Apache 2.0. | 6 | 18.5 | -0.04 | ||
| #516 | emotional-memory-agent mitemotional-memory-agent: A multi-tenant emotional memory system powered by an LLM agent, featuring user-level privacy isolation, long-term memory persistence, RAG retrieval, diary writing, and support for personalized interactions in QQ private and group chats. | 3 | 18.5 | 0.00 | ||
| #532 | claude-soul mitmcp-serverclaude-soul: Self-correcting learning engine for Claude Code — persistent identity, behavioral pattern tracking, and cross-session memory. | 6 | 18.2 | 0.00 | ||
| #539 | jayoohwang1/webchallenger ide-pluginjayoohwang1/webchallenger: Autonomous web navigation remains challenging for LLM agents, and the strongest generalist systems rely on proprietary reasoning models whose inference cost is prohibitive for the repetitive tasks where such agents would be most useful. We argue this gap stems not from insuffi... | 8 | 18.1 | 0.00 | ||
| #545 | llm-agent-memory-gateway llm-agent-memory-gateway: Standalone module split from costa92/llm-agent-ecosystem umbrella (llm-agent-memory-gateway). | 9 | 18.0 | 0.00 | ||
| #544 | llm-agent-memory-contract llm-agent-memory-contract: Standalone module split from costa92/llm-agent-ecosystem umbrella (llm-agent-memory-contract). | 9 | 18.0 | 0.00 | ||
| #546 | llm-agent-memory-postgres llm-agent-memory-postgres: Standalone module split from costa92/llm-agent-ecosystem umbrella (llm-agent-memory-postgres). | 9 | 18.0 | 0.00 | ||
| #547 | llm-agent-memory-worker llm-agent-memory-worker: Standalone module split from costa92/llm-agent-ecosystem umbrella (llm-agent-memory-worker). | 9 | 18.0 | 0.00 | ||
| #578 | @miller-tech/uap ide-plugin@miller-tech/uap: Autonomous AI agent memory system with CLAUDE.md protocol enforcement | 2 | 17.3 | 0.00 | ||
| #595 | FERNme mcp-serverFERNme: A lightweight memory engine for AI agents using fuzzy graphs, Hebbian updates, and optional LLM gating. | 116 | 17.0 | +1.99 | ||
| #603 | shokunin mitmcp-servershokunin: 職人 Shokunin — 62 AI agent skills for OpenCode, Claude Code, Cursor, Windsurf. ChromaDB memory, MCP servers, declarative self-updates. Multi-model, open source, zero cost. | 4 | 16.9 | -0.02 | ||
| #606 | ai-shared-brain mitide-pluginai-shared-brain: Give your AI agent persistent memory with markdown files. Starter kit from the video: My AI and I Share a Brain. | 3 | 16.8 | 0.00 | ||
| #619 | cometline-release apache-2.0cometline-release: AI companion for your workspace, with per‑project memory, coding agents, and a native desktop chat UI. | 7 | 16.6 | 0.00 | ||
| #618 | cometline apache-2.0cometline: AI companion for your workspace, with per‑project memory, coding agents, and a native desktop chat UI. | 7 | 16.6 | 0.00 | ||
| #629 | llm-agent-orchestrator mitlibraryllm-agent-orchestrator: A high-performance multi-agent framework with stateful task routing, shared semantic memory, and tool integration. | 6 | 16.5 | 0.00 | ||
| #626 | iDEA-iSAIL-Lab-UIUC/TimeClaw ide-pluginiDEA-iSAIL-Lab-UIUC/TimeClaw: Time series are often embedded in rich contexts that are essential for holistic modeling. Moreover, real-world practitioners often require end-to-end workflows for analyzing temporal dynamics, where widely studied tasks such as forecasting are only one step in a broader soluti... | 6 | 16.5 | 0.00 | ||
| #659 | agent-memory apache-2.0agent-memory: Verifiable memory for LLM agents — every recalled claim is HMAC-signed back to its originating trajectory span. | 9 | 16.2 | 0.00 | ||
| #709 | bhived-mcp apache-2.0mcp-serverbhived-mcp: bhived is an MCP server that gives AI agents shared memory, skills, and tool discovery. install once, works in Claude Code, Cursor, and 15+ other agents. | 15 | 15.1 | -0.15 | ||
| #753 | mnemo mitmcp-servermnemo: Persistent memory for Claude Code. Your AI never starts from scratch again. | 20 | 14.5 | 0.00 | ||
| #756 | LanceZPF/agent-as-a-router ide-pluginLanceZPF/agent-as-a-router: Real-world users typically have access to multiple Large Language Models (LLMs) from different providers, and these LLMs often excel at distinct domains, yet none dominate all. Consequently, routing each task to the most suitable model becomes critical for both performance and... | 166 | 14.5 | +3.78 | ||
| #797 | @loongsuite/opentelemetry-util-genai @loongsuite/opentelemetry-util-genai: OpenTelemetry GenAI utility for standardized telemetry collection across LLM, Agent, Embedding, Tool, Retrieval, Rerank, Memory and more. | 38 | 13.7 | +1.04 | ||
| #798 | iva mitmcp-serveriva: Open-source personal AI agent with long-term memory. The best tools, hand-picked and assembled — one command and it works. | NEW | 13.7 | — | ||
| #800 | OneBrain mitmcp-serverOneBrain: Unify every AI agent on ONE local shared memory over MCP. Obsidian vault + gbrain (by Garry Tan) + Ollama. 100% offline, $0 API. Orchestrate with OpenClaw or Hermes. | 43 | 13.7 | -0.35 | ||
| #804 | centaur-loop mitcentaur-loop: 半人马环 Centaur Loop：面向 AI Agent 反馈闭环、人类治理和记忆复盘的开源工作台 / Human-governed AI feedback loop workbench. | 21 | 13.6 | +0.02 | ||
| #818 | GateMem mitGateMem: GateMem: a benchmark and evaluation toolkit for memory governance in multi-principal shared-memory LLM agents. | 17 | 13.5 | +0.15 | ||
| #824 | dual-mem apache-2.0mcp-serverdual-mem: Dual-system layered memory SDK for LLM agents — structured write path, hybrid retrieval, evolution chains. | NEW | 13.3 | — | ||
| #825 | Mklink-AI-Probe mitcliMklink-AI-Probe: MKLink AI Probe is an all-in-one embedded debugging CLI for Cortex-M microcontrollers. It bridges the MKLink/MicroLink hardware probe with AI agents (Claude, OpenAI) to enable natural-language driven firmware flashing, real-time waveform visualization, memory inspection,... | 13 | 13.3 | +0.16 | ||
| #828 | DataArcTech/Bayesian-Agent ide-pluginDataArcTech/Bayesian-Agent: LLM agents increasingly rely on external inference conditions: prompts, tools, memory, SOPs, skills, and harness feedback. These assets can improve task execution without changing model weights, but they are often revised by heuristic reflection or by reusing observed successe... | 13 | 13.2 | +0.12 | ||
| #851 | zjunlp/LightMem2 libraryzjunlp/LightMem2: As LLM agents are deployed in long-horizon sessions, context accumulation drives up inference costs. Existing approaches utilize text pruning or dynamic memory eviction to minimize token footprints; however, their unconstrained sequence mutations alter layouts, introducing pre... | 27 | 12.8 | -0.11 | ||
| #853 | @loongsuite/otel-util-genai @loongsuite/otel-util-genai: OpenTelemetry GenAI utility for standardized telemetry collection across LLM, Agent, Embedding, Tool, Retrieval, Rerank, Memory and more. | 25 | 12.8 | -0.08 | ||
| #887 | ai-memory-reader ai-memory-reader: The native macOS & iOS app for browsing AI agent memory files — Claude Code, OpenClaw, Codex, Cursor, Gemini. | 20 | 12.1 | -0.05 | ||
| #901 | dejavu apache-2.0mcp-serverdejavu: Memory that follows you across every AI tool. No cloud storage. No account required. Set it up once, use it everywhere. | 22 | 11.8 | -0.05 | ||
| #908 | awesome-agent-memory awesome-agent-memory: Paper List of Agent Memory Systems along Four Axes. | NEW | 11.6 | — | ||
| #911 | llm-harness libraryllm-harness: Production-grade reusable AI agent infrastructure base. ReAct loop, tool pipeline, permissions, hooks, session persistence, memory consolidation, observability — ~10K lines, MIT license. | 29 | 11.5 | -0.26 | ||
| #920 | ai-memory-context-management libraryai-memory-context-management: Cache-optimised architecture and Shadow VFS framework for token-efficient LLM agents. | 20 | 11.3 | +0.22 | ||
| #923 | @adminforth/agent @adminforth/agent: AI agent plugin for AdminForth with tool-based workflows and persistent chat sessions. | 52 | 11.2 | -0.87 | ||
| #926 | agd-memory mitmcp-serveragd-memory: Per-project persistent memory for Claude Code, in AGD format. Skill + SessionStart hook + MCP server, all driven by the agd CLI. | 28 | 11.1 | 0.00 | ||
| #975 | llm-agent-memory-management llm-agent-memory-management: discovered AI agent. | 37 | 10.2 | -0.14 | ||
| #978 | RecMem mitRecMem: This repository contains the implementation of the paper "RecMem: Recurrence-based Memory Consolidation for Efficient and Effective Long-Running LLM Agents". | 29 | 10.2 | 0.00 | ||
| #989 | PRISM mitide-pluginPRISM: A training-free, retrieval-side memory framework for long-horizon LLM agents — high accuracy at an order-of-magnitude smaller context budget. | 29 | 10.0 | 0.00 | ||
| #1039 | llm-agent-tools ide-pluginllm-agent-tools: Lightweight LLM agent library — configurable memory layers, skills, context engineering. Works with any LLM provider | 35 | 9.1 | +0.01 | ||
| #1064 | antman9914/proplay ide-pluginantman9914/proplay: Self-evolving agents are expected to improve through interaction without external supervision, but this remains difficult in partially observable environments where agents must explore actively, learn from limited feedback, and decide when to trust prior experience. Existing L... | 6 | 8.7 | +0.58 | ||
| #1065 | awesome-llm-agent-skills-papers awesome-llm-agent-skills-papers: A curated list of papers, blog posts, and systems on skills for LLM agents — reusable, named capability units that an agent can store, retrieve, compose, and improve over time — together with closely adjacent research on tool use, function calling, procedural memory, and skill... | 38 | 8.7 | 0.00 | ||
| #1068 | neuromem_runtime mitneuromem_runtime: Local-first memory for long-running LLM agents. | 36 | 8.6 | 0.00 | ||
| #1130 | progressive-synaptic-memory mitprogressive-synaptic-memory: Progressive Synaptic Memory: a token-efficient long-term memory mechanism for LLM Agents / 渐进式突触记忆机制. | 35 | 7.5 | 0.00 | ||
| #1142 | llm-agent-memory llm-agent-memory: Multi-service memory platform for the llm-agent ecosystem (SDK + Gateway + Postgres + workers). | 34 | 7.4 | 0.00 | ||
| #1150 | role2026/rolememo ide-pluginrole2026/rolememo: While role-playing agents excel in short-term interactions, long-term conversations overwhelm context windows, motivating external memory frameworks. Current systems typically rely on persona-agnostic summarization, which records facts without persona-specific interpretation,... | 32 | 7.2 | 0.00 | ||
| #1152 | @alekra1/llm-wiki ide-plugin@alekra1/llm-wiki: Markdown wiki template for LLM agent persistent memory. Gives agents cross-session, cross-provider memory. | 54 | 7.1 | -0.33 | ||
| #1180 | llm-naming-game mitllm-naming-game: Networked naming game with bounded-memory LLM agents. Simulation code for Mehdizadeh & Hilbert (2026). | 33 | 6.5 | 0.00 |
Browse all sectors at /sectors.