CrewAI + MemorySync

Overview

Install crewai-memorysync and point Memory(storage=MemorySyncStorage(...)) at MemorySync: CrewAI’s own save/recall loop runs against durable server rows instead of local LanceDB files, so crew memory survives redeploys and is shared across services. Five agent tools and a one-line task-context helper complete the surface. The only memory vendor implementing CrewAI 1.x’s actual StorageBackend protocol.

Setup status and requirements

Supported version

crewai-memorysync 1.0.0 (PyPI)

Last setup review

2026-08-23

Permissions

A MemorySync API key with read and write scopes; deletes and resets need delete permission.

Limits

Requires crewai 1.x (Python 3.10+). Full resets require allow_full_reset=True by design. CrewAI embeds queries with its own embedder; keep one embedding model per store.

Capabilities

• Genuine StorageBackend: CrewAI drives saves and recalls automatically
• Memory survives redeploys and is shared across crews and services
• Built-in-parity search: same score scale and filter semantics as LanceDB
• Retry-safe idempotent writes; failures are loud, never silently dropped
• Full-wipe guard: reset(None) refused without an explicit opt-in
• Five never-raise agent tools with a read-only mode
• One-line register_memorysync() for Memory(storage="memorysync")

Quick Start

from crewai import Crew
from crewai.memory import Memory
from crewai_memorysync import MemorySyncStorage

crew = Crew(
    agents=[researcher, writer],
    tasks=[research_task, write_task],
    memory=Memory(storage=MemorySyncStorage(user_id="support-crew")),
)
crew.kickoff()

Use Cases

Explore More