Guides¶

Comprehensive guides for SPADE_LLM features and concepts.

Core Concepts¶

Architecture - System components and design
Providers - LLM provider configuration and usage
Tools System - Function calling and tool integration
Structured Output - Type-safe structured responses with Pydantic schemas
RAG System - Retrieval-Augmented Generation with multi-agent collaboration
Context Management - Advanced context control and message management
Memory System - Dual memory architecture for agent learning and conversation continuity
Memory Architecture - Detailed memory system architecture and diagrams
Coordinator Agent - Organizational orchestration with shared context and routing
Conversations - Conversation lifecycle and management
MCP - Model context protocol integration
Human-in-the-Loop - Human expert consultation and workflows
Guardrails - Content filtering and safety controls
Routing - Message routing and multi-agent workflows

Each guide covers: - Core concepts and configuration - Common usage patterns - Best practices and troubleshooting - Complete code examples