phoenix
Level AMCP ecosystemNot an MCP ServerSourceAI Observability & Evaluation
Overview
Arize Phoenix is an open-source AI observability platform for experimenting, evaluating, and troubleshooting LLM applications. It traces runtime behavior with OpenTelemetry-based instrumentation, runs LLM-driven response and retrieval evaluations, manages versioned datasets, tracks experiments on prompt, model, and retrieval changes, and includes a playground for prompt optimization and model comparison. An MCP server lets MCP-compatible clients connect to Phoenix.
Installation
View source repo →Related servers
serverAI Observability & EvaluationNetdataReal-time infrastructure monitoring with metrics, logs, alerts, and ML-based anoToolJetOpen-source foundation of ToolJet AI - the enterprise app generation platform fosignozSigNoz is an open-source, OpenTelemetry-native observability platform for your tkuboard-pressKuboard v4 官方文档（zh/en 双语）：Kuboard 是基于 Kubernetes 的微服务管理界面。本仓库为 v4 版本文档站源码，涵盖安装升级SkillSpectorSecurity scanner for AI agent skills. Detect vulnerabilities, malicious patterns