Agent frameworks & infrastructure
Agent memory & context
4,032 agents · most starred first
Agent memory & context, page 36 of 41
Agentic Memory CLI — persistent memory for AI agents. As easy as git.
Memory system for LLM agents -- extraction, retrieval, hybrid search, and MMR reranking
Procedural memory for LLM agents -- skill definitions + invocation history
Long-running-agent foundation -- wake schedules, fires, webhook subscriptions
Context recovery for AI agents — analyze fatigue, compress context, improve token efficiency
Smart context compression for LLM agents — preserve critical info, reduce tokens 40-80%
Real-time memory health monitor and context rot detector for AI agents
Persistent semantic memory for AI agents with replay, restore, verify, and ignore decisions
Claude learns who you are from your own sessions and stays that way
Plain-text memory for AI agents — small, filtered, compressed, self-pruning. No database, no embeddings, no dependencies.
Framework-agnostic message reducer for AI agent state management. Works with LangGraph, CrewAI, and plain dicts.
Git for LLM context. Snapshot a dying session, brief the next model in 30 seconds.
Long-term memory tier for AI agents backed by Valkey Search: semantic recall with recency/importance ranking, scoped capacity eviction, and consolidation. Pairs with @betterdb/agent-cache.
AutoGen integration for CortexDB — long-term memory for AI agent conversations
LangChain integration for CortexDB — long-term memory for AI systems
Python SDK for the Couchbase Agent Memory API
Mimir long-term memory as CrewAI tools — local-first, encrypted, persistent memory the agent can explicitly call.
Azure CosmosDB persistence backend for CrewAI Flows with built-in message pruning via agentstate-reducer
Google Firestore persistence backend for CrewAI Flows with built-in message pruning via agentstate-reducer
Persistent, queryable knowledge storage for AI agents. Markdown source + SQLite + manifest fallback.
Drop-in LangGraph store that merges concurrent writes instead of silently dropping them.
Mimir persistent, local, encrypted memory for LangChain — tools and a retriever backed by the Mimir MCP engine.
ScyllaDB checkpointer for LangGraph
SurrealDB checkpoint persistence and cross-thread stores for LangGraph.
LangGraph checkpoint saver for Azure CosmosDB with built-in message history pruning via agentstate-reducer
LangGraph checkpoint saver for Google Firestore with built-in message history pruning via agentstate-reducer
Python SDK for OpenMemory and Hermes AI integration
Mimir persistent, local, encrypted memory for Pydantic AI agents — exposed as a Toolset.
Host-agnostic, self-correcting library/API memory daemon for coding agents (Qwen MemoryAgent)
Drop-in state backend for LangGraph, CrewAI & custom agent loops - Rust core, framework-agnostic, built for production.
LangGraph integration for Zep
Microsoft Agent Framework integration for Zep
Pydantic AI integration for Zep
Perseus Vault long-term memory as CrewAI tools — local-first, encrypted, persistent memory the agent can explicitly call.
Perseus Vault persistent, local, encrypted memory for LangChain — tools and a retriever backed by the Perseus Vault MCP engine.
Grafomem (Fair Source) adapter for LangGraph Checkpointers
The intelligent context management framework for LLM applications.
Perseus Vault persistent, local, encrypted memory for Pydantic AI agents — exposed as a Toolset.
Agents that remember: memory-first agent & RAG framework with zero required dependencies.
TypeScript SDK for MemoryMesh durable agent memory, receipts, checkpoints, and Cognee-backed runtime state.
LangGraph Checkpoint saver using native SQLite for Node.js and Bun.
Three-tier LangGraph checkpoint saver: L1 SQLite -> L2 NATS KV -> L3 PostgreSQL
⚡ B-FAST Context Compression Layer for Large Language Models (LLMs)
Zero-knowledge encrypted sync for AI agent context
Agentic MCP memory server with dream cycle — decay, consolidation, and archival
Task-aware memory management for LLMs. Extend conversations 5.7x longer with 88% recall.
Haystack integration for Mirage, a unified virtual filesystem for AI agents
LangGraph BaseStore backed by Sibyl Memory (SQLite + FTS5, no vector database, no embeddings).
Pool-split retrieval for agent memory: reserved per-type token budgets so low-priority entries surface and corrections never get crowded out.
An Enterprise AI Agent Optimization Platform & Context Intelligence Operating System
Keep your AI agent context small and cheap by managing tool bloat and large outputs.
Local-first mailbox for intentional AI agent context handoffs.
CLI and MCP server for OmniContext — AI agent project memory layer
Give your AI agent a memory. And a seatbelt.
cogmem — agent-native memory kernel for a single local-first AI agent
Context deduplication for LLM agent sessions: repeated tool output is replaced with a short ref token
Persistent self-improving agent memory with tree-of-thoughts reasoning: MCTS/beam search, PRM step verification, self-consistency voting, s1 budget forcing, REPL program-aided reasoning, reflexion self-correction, failure classification, experience replay
Rulvar PostgreSQL store implementing JournalStore and LeasableStore with a fencing epoch, for multi-process and multi-host deployments.
Manage human-curated context assets for LLMs and agents in Git.
정크 컨텍스트 제거 — LLM 에이전트를 위한 보이지 않는 정보 정돈 레이어
Rulvar SQLite store implementing JournalStore and LeasableStore with a fencing epoch.
In-memory SkeinStore + queue driver for development and tests.
Postgres SkeinStore driver with pgvector semantic search, reusing PostgresSaver for checkpoints.
Archil filesystem adapter for LangChain
LangChain.js integration for the Taguru long-term semantic memory server
Official LangChain JS / LangGraph JS adapters for Wolbarg shared memory — BaseMemory + BaseStore.
Auto-apply Grok 4.5 context compaction to xAI Responses API agent loops. Drop-in client that compacts every N turns or when a token budget is exceeded.
MongoDB-backed persistent memory provider for the Vercel AI SDK (plus a Vercel Eve provider that is not yet production-ready) — Session, Semantic, Procedural, Episodic, and Scratchpad memory with Atlas Vector Search (optional Atlas Automated Embedding).
Provider-neutral second-brain runtime and contract helpers for human and agent operators.
WebAssembly/JS bindings for the llmtrim-core compression engine (wasm-bindgen).
TormentNexus for OpenHands — Persistent AI memory, MCP tools, and context harvesting in your autonomous agent
Active state management for multi-agent coding: a shared, versioned project brain over MCP.
Canonical Memorix metadata contracts, effective multi-agent resolution, and metadata-only abstract unions (Phase 3).
Advanced memory management system for AI agents with semantic retrieval, hierarchical structures, and multi-agent support
Local-first persistent vector memory, MCP tooling, and multi-agent workflows for AI assistants
Multi-agent memory intelligence layer — shared knowledge, expertise tracking, and conflict detection for AI agent teams
File-backed Memory MCP Server for multi-agent coordination
CogX CLI for the Persistent Memory Protocol (PMP): durable AI identity, memory, and multi-agent orchestration.
The who's-who / identity / relationship substrate for agents — trust ladder (family/friend/acquaintance/stranger), multi-party and multi-agent, consumed via the FriendStore interface + FriendResolver.
Memory-native agent framework runtime.
myai — a self-owned, git-versioned AI memory + multi-agent framework. Plug any agent (Claude, Cursor, Gemini, Codex, opencode) into your own portable brain so it boots already knowing your projects, decisions, and context. Local-first (Docker), MCP-native
AIGNE AFS Memory Provider — agentic knowledge lifecycle management
CREWAI adapter for Context Router - brings structured state to CREWAI crews and agents
Compress a completed agent's trace into a compact, framework-agnostic briefing that the next agent can load in one call. Solves the cross-framework agent-to-agent handoff problem. Paid via x402 ($0.10/pack) on Base USDC.
CDS Plugin for LangGraph checkpoint persistence in SAP CAP applications
Evidence-preserving LangGraph Store adapter for Engram.
SurrealDB checkpoint saver, store and node cache for LangGraph.js, plus a SurrealDB Agent Memory store
A versioned, searchable JSON workspace for AI agents: scoped edits, semantic discovery, audit history, time-travel, and delta persistence.
Versioned identity blocks for LLM agents -- self-evolution via a linear version chain with propose/consent/apply, rationale, and rollback
Standing-wants corpus for LLM agents -- intention lifecycle, salience decay, embedding dedup, and private deliberation tools
Governed-knowledge retrieval + injection for LLM agents -- concepts, playbook entries, three-scope shadow merge, and before_model middleware
ACID-style transaction guarantees (rollback + session-memory guardrails) for autonomous AI agents
Local MCP server that saves work-in-progress checkpoints so any AI agent can resume a multi-step plan exactly where a previous session was cut off.
Core infrastructure for AI agents - memory, knowledge compilation, and tooling
Azure Cosmos DB Agent Memory Toolkit integration for Microsoft Agent Framework - semantic memory with fact extraction and user profiles.
Conversational memory as a knowledge graph
Local SQLite long-term memory for AI assistants, served over MCP
Local-first MCP memory backend and governance console for coding agents
Two-stage retrieval + gating long-term memory engine for coding agents.
Agent Memory File System - a local-first memory store for AI agents, with semantic search.