Memory

Configure the OpenClaw persistent memory system for long-term knowledge retention across conversations.

Overview

The Memory system allows agents to store and recall information across conversations. This enables agents to remember user preferences, past interactions, and accumulated knowledge.

How Memory Works

Short-term — Conversation history within a session
Long-term — Persistent facts stored across sessions
Shared — Knowledge shared across agents

Configuration

{
  "memory": {
    "enabled": true,
    "backend": "local",
    "maxEntries": 1000,
    "ttl": "30d"
  }
}

Options

Option	Description	Default
`enabled`	Enable memory system	`true`
`backend`	Storage backend (local, redis)	`local`
`maxEntries`	Maximum memory entries	`1000`
`ttl`	Time-to-live for entries	`30d`

Memory Management

# List stored memories
openclaw memory list

# Clear all memories
openclaw memory clear

# Export memories
openclaw memory export > memories.json

Per-Agent Memory

Each agent can have isolated memory:

{
  "agents": {
    "assistant": {
      "memory": {
        "scope": "per-sender",
        "maxEntries": 500
      }
    }
  }
}

Next Steps

Skills — Custom skill configuration
AI Agent — Agent configuration
Security — Security best practices