Let’s explore some friendly ways to break down those large system prompts into smaller, more manageable pieces: • Multi-Agent Workflows: Think about splitting complex tasks into smaller subtasks, with each handled by a dedicated agent using a focused system prompt. For instance, in a customer onboarding process, you might have separate agents taking care of user information collection, identity verification, payment processing, and enrollment confirmation. This way, everything feels organized and clear! • Hierarchical Chunking: When dealing with large text inputs, try breaking the content into manageable segments while keeping important structural markers intact. You can analyze each chunk on its own and then combine the results step-by-step to create a full, comprehensive output. Just to better assist you, are you working on a deterministic workflow like automation, or are you processing large text inputs? Knowing this will help me guide you toward the best approach tailored to your needs. Sources: Building the all-in-one platform for voice AI agents | Building the all-in-one platform for voice AI agents | Building the all-in-one platform for voice AI agents | Knowledge Management & Data Ingestion System | livekit/livekit_composite | ... | Knowledge Management System | livekit/livekit_composite | DeepWiki