Crate llm_edge_agent 

LLM Edge Agent - High-performance LLM Intercepting Proxy

This crate provides the main application logic for the LLM Edge Agent, integrating all layers into a complete end-to-end system:

• Layer 1: HTTP Server (Axum) with auth and rate limiting
• Layer 2: Multi-tier caching (L1 Moka + L2 Redis)
• Layer 2: Intelligent routing with circuit breakers
• Layer 3: Provider adapters (OpenAI, Anthropic)
• Cross-cutting: Observability (Prometheus, OpenTelemetry, Logging)

Re-exports

pub use integration::check_system_health;

pub use integration::initialize_app_state;

pub use integration::AppConfig;

pub use integration::AppState;

pub use proxy::handle_chat_completions;

pub use proxy::ChatCompletionRequest;

pub use proxy::ChatCompletionResponse;

Modules

integration

Integration module - orchestrates all system components

proxy

Proxy request handler