Integrations
Works with your stack
Connect the tools your team already uses. Each integration syncs permitted content into MemorySync so your AI agents retrieve it alongside their application memory.
AG2
An automatic memory loop for AG2’s ConversableAgent — one attach call, duplicate-proof multi-agent capture, and a deadlock-free bridge.
AgentOps
Memory operations as first-class spans in AgentOps: the first SDK-level tracing instrumentor from any memory vendor, wired through standard OpenTelemetry.
Agno
A memory-only db for Agno’s MemoryManager: automatic extraction and injection, real semantic recall, an async-native twin, and deletes that cannot nuke an account.
Amazon S3
Index documents from the S3 bucket paths you approve, including Cloudflare R2, MinIO, Wasabi, and Backblaze B2.
AutoGen
A real Memory-protocol implementation for Microsoft AutoGen — async-native, budgeted recall, and a clear() that cannot nuke a customer.
AWS Bedrock
A runnable reference architecture: a memory-augmented Converse loop plus the first Bedrock Agents action-group memory integration from any vendor.
CAMEL-AI
A lossless storage backend plus a drop-in AgentMemory: verbatim history, scored semantic recall under a hard budget, and zero client-side embeddings.
ChatDev
A memorysync store type for ChatDev’s zero-code platform: one pip install adds it to workflow YAML and the web console — semantic recall with budgets, honest timestamps, both-side capture.
Claude Code
Automatic memory for Claude Code and Claude Cowork: hooks capture every exchange and recall context per project.
Codex
Automatic memory for Codex: hooks inject recalled context and capture every exchange, TOML + bearer env auth.
CrewAI
The native CrewAI 1.x memory backend: durable, shared, cross-execution crew memory via Memory(storage=...).
Cursor
Automatic memory for Cursor: hooks capture every exchange, an always-on rule keeps recall running.
Devin (Windsurf)
Memory for Devin Desktop and the Devin CLI — MCP on current and legacy paths, first-vendor hooks.v1 wiring.
Dify
A first-party Tool plugin: recall context before LLM nodes, remember facts after them — the cross-conversation user memory Dify does not have natively.
ElevenLabs
Memory for ElevenLabs Agents: a Custom-LLM proxy with a hard recall budget, phone-caller identity, and HMAC-verified post-call capture.
Eve
The first installable memory package for Vercel’s Eve agent framework: three one-line mounts wire persistent, per-user memory into any Eve agent directory.
Flowise
A MemorySync Memory node with SEPARATE user and session scopes: long-term memory follows the user across chats while each chat keeps its own transcript.
GitHub
Sync codebases, pull requests, and issues directly into MemorySync to give your AI agents full engineering context.
Google ADK
A real BaseMemoryService whose ingestion actually stores sessions — powering ADK’s native memory tools.
Google Antigravity
A full Antigravity plugin: hooks inject recalled context and capture every exchange across AGY, the AGY IDE, and the AGY CLI.
Google Drive
Ingest Google Docs, Sheets, and Slides into your agent's memory graph with real-time continuous synchronization.
Granola
Index AI meeting notes, summaries, and transcripts from the Granola folders you approve, so what a meeting decided stays searchable.
Hermes Agent
A native Hermes memory provider: persistent profiles, background sync, and zero-latency prefetched recall — zero dependencies.
LangChain
Persistent chat history and prompt-ready recall context for LangChain chains and agents, in Python and Node.js.
LangChain Tools
Structured memory tools that let LangChain agents store, search, list, update, and delete memories safely.
LangGraph
Durable cross-thread memory for LangGraph agents: a native store, context injection, and turn persistence.
LiveKit
Voice agents that remember callers — recall injected under a hard latency budget, so a reply is never late.
LlamaIndex
LlamaIndex’s Memory with immediate durable persistence, native block recall, and memories as a RAG retriever.
Mastra
Memory on Mastra’s native processor pipeline: recall before each call, persistence after it, per-call identity.
Microsoft Agent Framework
A ContextProvider with budgeted instruction-layer recall, per-run identity, and capture that can never crash the agent.
Microsoft OneDrive
Sync files from a connected user's OneDrive and accessible SharePoint document libraries.
n8n
Two nodes: workflow memory operations, and the only maintained AI Agent memory sub-node from a memory vendor.
Notion
Transform your Notion workspace into a searchable, temporally-aware knowledge graph for your AI agents.
NVIDIA NeMo Agent Toolkit
A pip-installable MemoryEditor plugin: one YAML block gives any toolkit workflow long-term memory — explicit memory tools or fully automatic via auto_memory_agent.
OpenAI Agents SDK
The first drop-in Session implementation: durable history across restarts and handoffs, plus long-term memory.
OpenClaw
OpenClaw’s memory slot, owned by MemorySync: automatic recall and capture on every interaction, across every channel.
OpenCode
One plugin line: recalled context injected per session, every exchange captured, MCP and slash commands registered.
Pipecat
One FrameProcessor gives Pipecat pipelines budgeted memory recall and delta-only persistence — on the current frame API.
Pydantic AI
One capability gives every run recalled context and success-only turn persistence — no history pollution.
Respan
MemorySync memory + Respan (formerly Keywords AI) gateway observability: every recall, LLM call, and persist is a first-class traced step in cost-tracked applications.
Slack
Index messages and thread replies from the Slack channels you approve, so decisions made in chat stay searchable.
Strands Agents
A MemoryStore with automatic recall before every model call, turn-by-turn extraction, and a crash-proof lifecycle — zero framework dependencies.
Vercel AI SDK
Long-term memory for generateText and streamText with one wrapLanguageModel call, on AI SDK v6 and v7.
VS Code
Memory for GitHub Copilot in VS Code: one-click MCP install, exact configs for every scope, and the coding agent on github.com.
Web Crawler
Crawl and ingest permitted public web pages and documentation into MemorySync.
Zapier
Memory operations for 8,000-app automations — including the category’s first New Memory trigger.
Built for production AI systems
Build AI systems that remember
MemorySync provides the infrastructure layer for persistent memory, adaptive retrieval, and enterprise AI intelligence.
Start free. No credit card required.