Memory and State

Learn how IrisOS agents store and recall information across conversations.

Overview

IrisOS agents maintain memory and state to provide context-aware, personalized interactions. The memory system enables agents to remember past conversations, user preferences, and learned information.

Memory Types

Short-Term Memory

Conversation context within a single session:

{
  "memory": {
    "shortTermContext": 10,
    "contextWindow": "last_10_messages"
  }
}

Long-Term Memory

Persistent storage across sessions:

{
  "memory": {
    "enabled": true,
    "longTermEnabled": true,
    "rememberUsers": true,
    "rememberConversations": true,
    "retentionDays": 90
  }
}

Knowledge Base

Static information the agent knows:

{
  "knowledge": [
    "Expert in DeFi protocols",
    "Understanding of yield farming",
    "Knowledge of liquidity pools"
  ]
}

State Management

User State

// Store user preferences
await agent.state.setUserData(userId, {
  preferences: { language: 'en', theme: 'dark' },
  history: []
});

// Retrieve user state
const userData = await agent.state.getUserData(userId);

Conversation State

// Track conversation context
await agent.state.setContext(conversationId, {
  topic: 'DeFi',
  lastMessage: 'What is yield farming?',
  entities: ['DeFi', 'yield farming']
});

Memory Configuration

{
  "memory": {
    "provider": "postgresql",
    "cacheEnabled": true,
    "cacheTTL": 3600,
    "maxMemorySize": "100MB",
    "embeddings": {
      "enabled": true,
      "model": "text-embedding-ada-002"
    }
  }
}

Best Practices

1. Privacy: Respect user privacy, allow data deletion

2. Retention: Set appropriate data retention periods

3. Relevance: Store only relevant information

4. Performance: Use caching for frequently accessed data

5. Security: Encrypt sensitive information

Next Steps

• Runtime and Lifecycle

• Character Interface

Need Help?

• Discord Community

• GitHub Discussions