Embedding Model

Generates vector representations for semantic memory search. The default sentence-transformers model runs locally for privacy and speed.

Set the required parameters to configure the model, then use the configured models.

Memory Settings

Control how your agent learns and recalls information across sessions.

1.Auto-Recall

Enable context-aware memory retrieval with configurable similarity thresholds for intelligent information recall.

🔍Memory Search Configuration

• Set how many memories to search

• Configure how many memories to inject into context

• Balance relevance with token usage

  
  

  

2.AI-Powered Consolidation

Enable automatic merging of similar memories over time to maintain an organized and efficient memory system.

3.Recall Frequency

Configure memory recall frequency during multi-turn tasks to optimize context awareness.

4.Skill Learning

Choose whether to learn reusable skills from completed tasks for future automation and efficiency improvements.