Agent frameworks & infrastructure

Agent memory & context

4,032 agents · most starred first

Agent memory & context, page 12 of 41

Five-layer-memory-system★ 24

🧠 布洛陀·五层记忆系统 v2.2.10 · 一包多平台:WorkBuddy·Qoder CN·ZCode·Hermes·Zwork(AutoClaw) 全自动|豆包办公 一步粘贴|LobsterAI 适配收尾中暂不推荐|Trae Work/千问/TeleAgent 适配层保留待解锁 · macOS/Windows

mempalace-code★ 24

Offline-first AI memory for coding — 29 MCP tools, temporal knowledge graph, and local vector search. No API keys, no cloud, no server.

humanlike★ 24

Humanlike — deterministic persona, context, memory and privacy controls for conversational AI agents · Детерминированное поведение ИИ-агентов

memory-engineering★ 24

Practical patterns, starters & CLI tools for memory engineering with AI coding agents. Stop re-explaining the repo — design the memory. Get a Memory Ready score.

big-agentic-brain★ 24

Swarm intelligence for distributed AI agents. One agent learns, the whole swarm knows. Shared memories, reminders and agent-to-agent messages, ranked by trust. MCP and Claude Code ready.

MERIDIAN_Brain★ 24

Project Meridian by MattVidPro: give any AI chatbot or agent (Claude, ChatGPT, Gemini) a lasting memory and a consistent personality using plain Markdown files. No servers or coding needed.

Engram★ 23

Persistent RAG memory layer for AI agents - store observations, retrieve context by semantic similarity, self-prune on TTL

recall★ 23

AI memory layer that lives in your stack

slimcontext★ 23

Lightweight, model-agnostic chat history compression (trim + summarize) for AI assistants.

memx-core★ 23

Local-first personal memory & knowledge store for LLM agents. ローカルLLM向けの個人用メモリ/知識ストア(short / chronicle / memopedia / archive の4層構成)。

afu-brain★ 23

OpenClaw-compatible MASL safety gate with public RAG packs for memory-aware AI agents

ACE★ 23

基于 ACE (Agentic Context Engineering) 框架的 Python 实现,通过演进式上下文实现自我改进的大语言模型。

ContextGraph★ 23

Shared memory bus for MCP-compatible agents with permissions, subscriptions, and optional payments.

pi-memctx★ 23

Local-first memory context for Pi coding agents.

Connapse★ 23

Self-hosted knowledge base and memory for AI agents. MCP server and REST API with hybrid search, reranking and per-user permissions over your files, S3, Azure Blob, GitHub, Confluence and Jira. RAG on .NET 10 and pgvector.

awesome-agent-memory★ 23

Curated reading list & living survey on long-term memory for LLM agents — 989 papers, 529 archived PDFs, deep notes mapped to memory-kernel modules. Feeds the Ymem Research Radar.

null-drift★ 23
c0★ 23

An external memory for LLMs: a bi-temporal knowledge graph with hybrid (keyword + vector) retrieval and a self-improving reflection loop. Benchmarked to beat flat vector RAG on corrections and time-versioned queries.

agent-memory-wiki★ 23

Obsidian-powered daily logs, session logs, /prepforreset handoffs, and wiki janitor skills for AI agents.

spector★ 23

Memory backbone for AI agents. Four-tier store — working, episodic, semantic, procedural — with decay, consolidation, and association graphs. Fused semantic + SIMD-accelerated hybrid recall, in-process MCP, REST/gRPC, and language SDKs.

holographic-memory★ 23

Local-first associative memory and vector search for AI agents, with Rust, Python, Node.js, and MCP integrations.

engram★ 23

Self-hosted MCP server + dashboard giving Claude Code, Cursor & Hermes agents shared long-term memory over a git-backed markdown vault. Agent-native, Obsidian-compatible second brain — markdown RAG, no vector DB.

Daming-OS★ 23

A memory and continuous learning system designed exclusively for AI — your AI grows smarter the more you use it.

dsh-mnemon★ 23

Cross-agent, local-first persistent memory plugin for DeepSeek Harness (DSH), powered by Mnemon. It shares long-term memory across Mnemon-enabled agents and adds runtime memory, searchable project documents, semantic recall, knowledge graph, and a Sidebar UI.

dsh-continual-evolve★ 23

Continual self-evolution plugin for DeepSeek Harness: versioned, auditable, rollback-safe harness state refined from session trajectories, with a benchmark-driven validation loop.

portable-handoff★ 23

Maintain the same context between AI chats, models, and coding agents. Stop re-explaining your project every time you switch AI tools.

motif★ 23

The experience graph for AI agents. Self-hosted memory, human-reviewed decisions, and outcome-backed datasets for model evaluation and learning.

babyseat★ 23

OpenAI-compatible proxy and chat UI that runs over a distributed cloud-to-edge dynamic memory graph that keeps LLM agent context small: it injects relevant memories instead of the full chat history (−51% tokens on real traffic). Local, beta.

SRSA★ 22

The Spaced Repetition Systems for Agents

polyverse-agent★ 22

A state-based autonomous agent framework. Crafting digital entities with non-linear cognitive memory and self-evolving emotions. (Work in Progress)

mind-mem★ 22

Persistent AI memory for Claude Code, OpenClaw, and any MCP-compatible agent. BM25F + vector hybrid, governance-aware, local-first, zero-infrastructure.

harness-memory★ 22

Project memory layer for coding harnesses

CacheLane★ 22
Tree-Ring-Memory★ 22

Framework-agnostic, local-first memory lifecycle for AI agents: Rust CLI, SQLite/FTS recall, forgetting, audit, consolidation, DOX/Revolve adapters, and TUI.

CortexKG★ 22

Portable AI memory powered by knowledge graphs. Transform conversations into a persistent, explorable knowledge graph and give any LLM long-term, user-owned memory.

dsh-openviking★ 22

面向 DeepSeek Harness 的 OpenViking 检索、资源管理、自动召回与会话记忆插件

dsh-layered-memory★ 22

让 DeepSeek Harness 拥有跨会话长期记忆:AI 自动记住你是谁、你的项目和偏好,新会话直接带上背景,生活与工作记忆自动分开互不干扰,零配置无感运行 | Long-term memory for DeepSeek Harness: the AI remembers who you are, your projects and preferences across sessions, keeps personal and work memories separate, zero-config.

dsh-hermes-memory★ 22

Hermes-style persistent memory (MEMORY.md / USER.md) for DeepSeek Harness (DSH) — a faithful port of hermes-agent MemoryStore: bounded, file-backed, model-curated, frozen-snapshot injection.

membukkit★ 22

Long-term memory for LLM apps that shows its work: dated facts, as-of answers, supersession, and receipts on every answer. Library, CLI, local GUI, MCP.

Porrima★ 22

A self-hosted, memory-native AI agent framework and cross-platform GUI application.

jev-native-agent-with-extended-options★ 22

Research design for a Jev-native agent system:enable more options than jev provided with virtulization and paging, tool integration, external helper logits Top-k proposals with Jev-controlled fallback ,decision-aware hierarchical memory, and dependency-aware replanning.

ai-memory-hub★ 22

Shared memory & coordination layer for AI coding tools. Local-first, no cloud, no MCP required.

Memwyre★ 22

Memwyre is an open-source, universal memory infrastructure and persistent knowledge retrieval layer for Large Language Models (LLMs), AI agents, and custom applications.

dsh-chain★ 22

神经元记忆网络 让 DSH 记住「为什么这么决定」,而不只是结论。思维链记忆插件:8 个池分级遗忘、每轮注入、神经元 3D 星云浏览、夜巡自动整理。

minime-mcp★ 21

Universal infinite memory layer for Developer AI assistants. One shared brain across Claude, Cursor, Windsurf & more. 100% local, built on MCP standard. Stop re-explaining context

memcortex★ 21

Memcortex is a lightweight semantic memory layer for LLMs using open source embedding model and Weaviate for vector storage and retrieval.

throughline★ 21

Source-linked project history for AI work across nine tools. Local PostgreSQL, one-click processing, and per-purpose local, API or CLI models. English/German UI.

kabot★ 21

Next-Gen Autonomous AI Engineer. Ditch the stateless chatbots. Kabot brings Stateful Hybrid Memory, zero-hallucination workflows, and persistent multi-agent orchestration to your local machine. 🦊🚀🛡️

chonkify★ 21

Extractive document compression for RAG and agent pipelines. +69% vs LLMLingua, +175% vs LLMLingua2 on information recovery. Compiled wheels, try it out.

oc-mnemoria★ 21

Persistent shared memory (hive mind) for OpenCode agents, powered by the mnemoria Rust engine

remem★ 21

Persistent, queryable memory for AI agents with intake policy, smart recall, context packs, graph recall, and multi-agent scoping

engram★ 21

Geometric agent memory for Agents: 8-tool lean contract, harness injection at wake, session handoff. 79 MCP tools. Not a vector DB. Rust + MCP.

continuum★ 21

Your whole desktop context — everything you've seen, read, typed, and heard — as a single MCP your agent can use to understand you and how you work. Open source, local-first, on-device.

abmind★ 21

Persistent memory & RAG engine for AI agents — SQLite + FTS5 + embeddings, multi-stage recall, sleep maintenance, multi-host hooks

mem0-aio★ 21

Unraid CA template and Docker build for Mem0 (OpenMemory). This mega-container utilizes s6-overlay to seamlessly bundle the pre-compiled Qdrant Vector Database, the Python FastAPI/MCP server, and the Next.js Dashboard UI into a single, click-and-play local AI memory layer. Built for homelabs, privacy, and true agentic AI memory retention.

skill-everything★ 21

Git-versioned agent memory: agents that never make the same mistake twice. Anthropic-Skill folder standard, multi-runtime (Claude Code, Cursor, Gemini CLI, OpenCode).

apricity★ 21

Apricity — a fast, local-first web UI for MemPalace: browse, search, edit & curate your AI agent's memory, knowledge graph & vector store. Zero-dependency, self-hosted, MCP-friendly.

claude-code-second-brain★ 21

Turn Claude Code into an assistant with persistent memory: one interview-driven prompt builds your context files, rules and decision log, and every session starts already knowing you.

distil★ 21

Compression with a quality contract — cache-aware, causally-pruned LLM context compression for agentic runtimes, certified non-inferior across 7 domains. Works with any SDK.

zeromem★ 21

Tokenless memory system for Hermes Agent. Keeps every message word for word in one SQLite file and searches it with no LLM calls. Rust implementation of Zero-Mem (arXiv:2607.29377).

aiduMEI★ 21

aiduMEI·爱嘟优忆思——aidu Memory Engine Insight · 智能体通用智慧引擎——Excellent memory system for all agents.

asp-mem★ 21

Lightweight, provider-agnostic long-term memory for conversational agents

hafiza-os★ 21

Linux, macOS ve Windows için kaynaklı ikinci beyin. Claude Code, Codex ve Antigravity adaptörleri; yerel Markdown kasa, ayrı hafıza incelemesi, isteğe bağlı Mem0/Jev.

atlas-context★ 21

Help your agent understand your life, on your terms. Local photo clues, your explanations, reviewed context and read-only MCP. Experimental alpha.

pi-memory-system★ 21

pi 的持久记忆扩展:自动把会话中的结论、偏好与踩坑提取成原子记忆,归并进场景块并生成用户画像,后续轮次按语义召回注入。

user-behavior-personalization★ 21

Review preferences from selected conversations, apply them by task, and compare real AI responses. 本地双语工作台:语义学习、条件偏好与真实回答对比。

sediment★ 21

Self-hosted evidence store for coding agents.

pi-jarvis★ 21

Memory you can edit. Watchers that wake an idle Pi agent without model polling. A persistent side conversation, searchable history, optional archive encryption, native MCP and explicit controls.

daimon★ 21

Dream-briefing for AI coding agents: a cognitive checkpoint at session end, a 'while you were away' briefing at session start.

chatstack★ 20

Minimalist Context Management for message-based GPTs

its-over-9k★ 20

Persistent memory and agent lifecycle for Claude Code — because sessions shouldn't start from zero.

memory-viewer★ 20

Memory Viewer for OpenClaw AI Agents - Beautiful web UI for browsing and editing agent memory files

openmemory-plus★ 20

🧠 AI Agent Memory Management Framework - Dual-layer memory architecture with smart classification and automatic extraction

Autodidact★ 20

Self-learning AI agent that gets smarter and cheaper over time. Routes between local and cloud LLMs, learns from every interaction, remembers everything.

open-webui-memory★ 20

Automatically identify, retrieve and store memories from user conversations in Open WebUI.

palaia★ 20

Self-hosted hub that gives all your AI tools one shared memory, one toolbox, and a way to hand work to each other. Claude, ChatGPT, Codex, Gemini and any MCP client connect once; your data stays in plain Markdown files on hardware you control. Runs on a NAS, a Pi, or your own server.

vir★ 20

Karpathy's LLM Wiki, built from your Claude Code sessions: a folder of markdown you own, Obsidian-native when you want it. Keeps what you decided, learns rules from what you repeat, readable by any agent.

octobrain★ 20

Persistent memory for AI assistants — semantic search, knowledge graph, and MCP server integration. Built in Rust.

cerefox★ 20

Shared memory for your AI agents: what one agent learns, every other agent can find. Hybrid search, versioning and audit, over MCP, CLI and HTTP. Supabase or local Docker.

context-diamond★ 20

Auditable context capsules for LLM handoffs, coding agents, and OpenCode MCP workflows.

Hermes-Codex-Plugin★ 20

Local-first memory, recall, and skill evolution plugin for Codex agents.

Holix★ 20

Helix is ​​a powerful self-learning AI agent with memory, skills, and the ability to recall tools. It learns from successful tasks and creates reusable skills for the future.

neurodock★ 20

A local-first cognitive substrate for neurodivergent professionals. Gives Claude memory, a sense of time, a translator for corporate ambiguity, and a guardrail that refuses to amplify rumination, hyperfocus, or sycophancy. MCP-native. No telemetry. AGPL-3.0-or-later. Self-ID sufficient — no diagnosis gating.

memory-forest★ 20

A verifiable local memory architecture for long-running AI agents

MindCache★ 20

Agentic Memory System for Long-Term AI

boring-agent-memory★ 20

Canonical memory infrastructure: BM25 recall over trusted files for AI agents.

proofpress★ 20

The judgment layer for agents. Capture consequential judgments and turn hard-won expertise into better calls across agents and teams.

lethe★ 20

The best-benchmarked open-source AI memory system and the first AI memory built to forget.

Habitus-AI★ 20

Habitus-AI is fundamentally new approach to AI Agent Memory Systems and Cognitive Harness. Dual-Cypher system runs along perpendicular graphs. Mutate edges in real time. Builds habits and dynamic personality traits. Can call tool reflectively, connect complex concepts to learned words, and send plain text outputs without needing an LLM

context-mem★ 20

Your AI forgets everything between sessions. This fixes that — 98%+ retrieval accuracy, 100% on LongMemEval, 99% token savings. 44 MCP tools. Fully local, zero cost.

jev-agent-design-with-topk-logits-choices★ 20

Research design for a Jev-native agent system: tool integration, speculative parameter proposals, external helper logits Top-k proposals with Jev-controlled fallback ,decision-aware hierarchical memory, and dependency-aware replanning.Feature:Jev naturallanguage conversation prototype using external helper logits and dynamic Top-k token selection.

Recollect★ 20

Self-hosted AI memory and MCP coordination for coding agents. Persistent context across repositories and sessions, with knowledge graphs, hybrid search, and a desktop web UI. Built in Rust for Codex and Claude Code.

gitmemory★ 20

Git-versioned, contiguity-checked memory for AI agents: verbatim transcripts, byte-offset recall, no LLM

AI_Articles★ 20

Reference patterns and implementation guides for using multiple AI agents (Hermes, Claude Code, Minimax M3, OpenCode) in a shared knowledge commons. Each article ships with a portable pattern and concrete setup.

mnesis★ 19

A Python library that solves context window degradation in long-running LLM agents by moving memory management out of the model layer and into a deterministic engine.

mnemostroma★ 19

Offline RAM-first cognitive leer/coprocessor for AI agents and robotics. Solves "Context Abandonment" with 20-80ms latency using a dual-thread biomimetic memory architecture (ONNX + SQLite WAL).

cortex★ 19

Archived Oct 2026. Public snapshot of CORTEX v2.4. CORTEX now ships inside ATERNA agents: aterna.ai/cortex

ledgermind★ 19

Local-first persistent memory for AI agents — captures conversations, extracts durable knowledge, resolves conflicts, and reuses relevant context across agent clients.

codemem★ 19

A standalone Rust memory engine for AI coding assistants. Single binary, zero runtime deps. Graph-vector hybrid with MCP tools.

pi-lcm★ 19

Lossless Context Management for Pi. DAG-based summarization with full history recovery.

More in Agent frameworks & infrastructure