Agent frameworks & infrastructure
Agent memory & context
4,032 agents · most starred first
Agent memory & context, page 39 of 41
Reusable cross-conversation profile memory for DlightRAG
LangGraph checkpoint saver backed by local filesystem, GCS, or S3 via fsspec
MemorySync memory for LlamaIndex: a Memory subclass with durable server-side long-term memory, a recall memory block, a retriever, and agent memory tools.
Local-first layered memory and autonomous learning engine
Archive local AI chat histories to S3, freeing up disk space while keeping conversations searchable
The lifeforce kit: give any agent an identity, a home floor, a memory, a governed mind, and a diary — and it appears in the universe.
MemorySync memory for Pydantic AI: a drop-in agent capability that injects recalled context every run and sends each completed run's user prompt to fact extraction, plus typed memory tools and async helpers.
Self-improving skills for Claude Code — your agent learns, recalls, reinforces, and forgets. SQLite + FTS5 + local embeddings + MCP.
Python SDK for TollMeshCache - Distributed CRDT-based caching
Agent Resurrect SDK
Renamed: this package is now entroptics-context.
Private local hybrid text and visual memory provider for Hermes Agent
One SDK for MemStrata conversational memory and agent governance hooks
Autonomous Zero-Vector Memory Framework for AI Agents with Biological REM Consolidation (Gemini, Anthropic, OpenAI-compatible, xAI)
Memory tools and middleware for Microsoft Agent Framework with Smaran
Memory tools and middleware for OpenAI function calling with Smaran
Cadet Brainstem — an orchestration + measurement layer that reduces AI-agent context consumption.
Abstract compaction service seam (ctx.compaction) for the DeepSeek Harness
Cross-session memory for coding agents: plain files, git-native, no server
Boot-time integrity check for AI agent memory & convention files — catch silent convention loss, injection dilution, broken rollback anchors.
Four small, dependency-free memory and safety patterns for LLM agents: scratchpad state, fuzzy recall, pre-compaction checkpointing, and permission boundaries.
Deterministic ACID transaction and rollback engine for AI agents in <1µs
Persistent work state and minimal context for long-running AI agents
Standalone brain + knowledge-base kernel (agenti ecosystem)
Plug-and-play topological memory substrate for LLMs, LangChain, and autonomous agent swarms.
Ephemeral two-party context bridge. A pours context, gets a share link; B answers once on the page; A reads the merged exchange; content self-destructs (1h unresolved free). Free tier 3 melds/hour; $3.33 one-time extends a meld to 24h for async handoffs. HTTP API, MCP server, A2A card.
Ephemeral context URL — puts the context on a URL so neither side has to paste the block. Then the URL dies.
Screenpipe review: compare features, pricing, use cases, access model, and alternatives for this Memory agent in 2026. Searchable computer history for people and AI assistants.
Local-first AI Agent Memory — inspection and management dashboard
Node filesystem-backed AGENTS.md context section for ai-agent-sdk
Git for AI agent conversations.
Persistent agent identity for AI agents — a hand-edited identity layer, nightly-distilled experience, and lessons injected every session, shared across your whole fleet. Works with Hermes, OpenClaw, Claude Code, Pi, and opencode.
Local Markdown project memory for Codex, Claude Code, and opencode, with hook-based context and agent-reviewed knowledge.
Tembory Agent Memory for n8n AI Agents
LangChain.js adapter for token-budget: BaseMessage[] conversion and a TokenBudgetMemory class — context window and token budget management for LangChain agents.
Universal memory system for AI agents - CLI + MCP server + library API
Compress LLM tool output, reduce context tokens, and recover lossy originals from a local stash.
A TypeScript library for managing AI conversation state
Yuanyi (元忆) — boundary-aware, file-based memory for AI agents, packaged as a YottaMeta Agent Plugin.
Cross-platform CLI and MCP cognitive runtime for AI coding assistants
A deterministic context compiler for AI agents — evidence-driven context engineering beside your markdown knowledge base: agents propose, evidence gates, you decide.
A shared record for AI agents: decisions, completed work, and commitments, with sources. Git-native engineering memory and a self-hosted state server.
Upstash AgentKit adapter for the Vercel AI SDK: chat history, memory tools, Redis-Search tools, rate limiting, and cached tools.
Unified access layer for RDS ContextDatabase: `ctxdb` CLI (memory + KB ops), one-shot multi-agent installer, per-agent config, hooks/plugins, and SKILL.md.
AI context layer and multi-agent coordination center for any repository, exposed over MCP.
Long-term memory for agentic tasks. It tells you when it doesn't know.
DynamoDB checkpoints for LangGraph with versioned channel storage, compression, optional S3 offload, and explicit recovery errors.
CDS Plugin for LangGraph long-term memory stores in SAP CAP applications
Hierarchical-graph long-term memory for AI agents: variable-depth abstraction tree, cross-topic recall, temporal knowledge graph. Ships an MCP server, an HTTP service, and a LangGraph BaseStore.
Deterministic archive and backup service for AI interaction histories.
Give your agent a filing cabinet.
Slim public Memory facade and models for Atlas Agent Engine
Message the CLI agents on your machine from your phone — every message becomes part of their memory.
Agent Memory Integrity (agmi): a conformance test suite that measures whether AI agent memory and checkpoint stores notice tampering
Local memory, a codebase index, and working discipline for AI coding agents. Zero dependencies, no network.
AgentScope long-term memory middleware backed by Agentbase Memory
Agentevo local runtime (TZ-2): capture, DLP, encrypted vault, recall, guards, execution.
The operating memory and learning runtime for AI agents — turn every execution into reusable intelligence.
Give AI agents inspectable observations of files, images, screens, and context with receipt-backed perception.
Shared core for CrewAI unified-memory StorageBackend implementations — build a Memory backend with five small primitives
Azure Cosmos DB StorageBackend for CrewAI unified Memory — hierarchical scopes, categories, metadata filters and native vector search (VectorDistance)
Amazon DynamoDB StorageBackend for CrewAI unified Memory — hierarchical scopes, categories, metadata filters and native vector search (SearchVectors)
Google Firestore StorageBackend for CrewAI unified Memory — hierarchical scopes, categories, metadata filters and native vector search (find_nearest)
PostgreSQL (pgvector) StorageBackend for CrewAI unified Memory — hierarchical scopes, categories, metadata filters and native vector search
Amazon DynamoDB persistence backend for CrewAI Flows with built-in message pruning via agentstate-reducer
MongoDB persistence backend for CrewAI Flows with built-in message pruning via agentstate-reducer
SQL persistence backend for CrewAI Flows via SQLAlchemy (SQLite, PostgreSQL, MySQL) with built-in message pruning via agentstate-reducer
Graph-backed context scratchpad for LangChain Deep Agents.
Scope-aware short-term, working, and long-term memory for HarborRAG.
GridGain checkpointer, store and node cache for LangGraph — short-term thread state, long-term cross-thread memory, and node-level caching.
LangGraph checkpoint savers on plain Redis (sync + async) - no RedisJSON / RediSearch modules required, works on Redis 5+
Long-term memory engine for LangGraph over any BaseStore — LLM fact extraction, consolidation, and recall ranked by semantic similarity, recency and importance (a maintained alternative to LangMem)
SingleStore checkpoint saver and store for LangGraph
Shared core for LangGraph long-term-memory (BaseStore) backends — implement a store with four small primitives
Azure CosmosDB long-term-memory store (BaseStore) for LangGraph — namespaced key/value memory with filters, list_namespaces and native semantic (Cosmos DB vector) search
Google Firestore long-term-memory store (BaseStore) for LangGraph — namespaced key/value memory with filters, list_namespaces and native semantic (Firestore vector) search
PostgreSQL long-term-memory store (BaseStore) for LangGraph — namespaced key/value memory with filters, list_namespaces and native semantic (pgvector) search
Lightweight semantic caching for LLM applications, backed by Redis.
Context assembly and memory for llmgrid
Amazon Bedrock AgentCore Runtime adapter for Memanto persistent memory (preview)
Version control for AI agent memory. Python SDK for Mnemosyne.
LangGraph BaseStore backed by Mnemosyne: versioned agent memory with blame and branch.
Pluggable open-source agentic memory layer — MRAgent (arXiv:2606.06036)
Self-hosted MCP memory for AI agents with evidence-backed, explainable claims
Azure Cosmos DB MemoryStore for the PydanticAI harness Memory capability — CAS versions via ETags and idempotent operation receipts
Amazon DynamoDB MemoryStore for the PydanticAI harness Memory capability — CAS versions and idempotent operation receipts via conditional writes
Shared core for PydanticAI harness MemoryStore backends — CAS versions, idempotent operation receipts and bounded search over a few conditional primitives
Pydantic AI Harness MemoryStore and Pixeltable catalog tools.
PostgreSQL MemoryStore for the PydanticAI harness Memory capability — the upstream transactional store, built from a connection URL with an owned asyncpg pool
Resume LangGraph agents from their own records, backed by Recall by Polign
Slim, fast, persistent memory + retrieval for LLM apps. numpy + httpx, ~1000 LOC, drops into anything.
Portable, auditable, cross-vendor memory for LLM agents: evidence-bound claims, trust tiers, and compiled context views.
Microsoft Agent Framework adapter for Taisce: a context provider that injects governed memory as one untrusted user message and records the turn afterwards.
LangGraph adapter for Taisce: agent middleware that injects governed memory as one untrusted human message and records the turn afterwards.
TESSERA — Temporal Evolving State Synthesis with Explicit Relations and Atomic Memories: a text-first, auditable memory and evidence layer for AI agents.
Social memory for multi-agent systems: who remembers what, about whom, visible to whom.
Sync AI agent sessions across machines via a git repository
Stash and resume AI agent sessions across Claude Code, Codex, opencode, and more.
TypeScript SDK for Centient — AI agent memory and context engineering infrastructure
Bitmap-frame context compression for vision-capable LLMs