llm-sentinel-api

REST API server with health checks, metrics, and query endpoints for LLM-Sentinel.

Overview

Production-ready REST API built with Axum:

• Health Endpoints: Liveness and readiness probes
• Metrics: Prometheus metrics exporter
• Query API: Historical telemetry and anomaly queries
• CORS: Configurable cross-origin support

Features

• Sub-10ms request latency
• Prometheus metrics integration
• Health check endpoints for Kubernetes
• Query API for historical data
• Request validation and error handling
• Rate limiting support

Usage

[dependencies]
llm-llm-sentinel-api = "0.1.0"

Example

use llm_sentinel_api::{ApiServer, ApiConfig};

#[tokio::main]
async fn main() -> Result<(), Box<dyn std::error::Error>> {
    let config = ApiConfig {
        bind_addr: "0.0.0.0:8080".parse()?,
        enable_cors: true,
        ..Default::default()
    };

    let server = ApiServer::new(config);
    server.start().await?;

    Ok(())
}

Endpoints

• GET /health/live - Liveness probe
• GET /health/ready - Readiness probe
• GET /metrics - Prometheus metrics
• GET /api/v1/telemetry - Query telemetry
• GET /api/v1/anomalies - Query anomalies

License

Apache-2.0