phoenix banner

English · 简体中文

Add Arize Phoenix MCP server to Cursor

Arize Phoenix 是 Arize 推出的开源 AI 可观测性（AI Observability）平台，专为实验对比、效果评估与故障排查而设计。针对托管式生产级工作流，Arize 亦提供 [Arize AX](https://arize.com/products/ax/)。Phoenix 核心功能包括： - [**_链路追踪 (Tracing)_**](https://arize.com/docs/phoenix/tracing/llm-traces) - 基于 OpenTelemetry 标准的插桩，追踪大语言模型（LLM）应用的运行时调用链。 - [**_效果评估 (Evaluation)_**](https://arize.com/docs/phoenix/evaluation/llm-evals) - 利用 LLM 自主能力，通过回答质量评估与检索效果评测建立基准测试。 - [**_数据集 (Datasets)_**](https://arize.com/docs/phoenix/datasets-and-experiments/overview-datasets) - 构建具有版本控制的示例数据集，用于科学实验、效果评估与模型微调。 - [**_实验对比 (Experiments)_**](https://arize.com/docs/phoenix/datasets-and-experiments/overview-datasets#experiments) - 追踪并评估提示词（Prompts）、底层模型（LLMs）及检索策略的演进变更。 - [**_试验场 (Playground)_**](https://arize.com/docs/phoenix/prompt-engineering/overview-prompts) - 优化提示词、横向对比不同模型、灵活调整参数，并一键重放历史追踪的 LLM 调用。 - [**_提示词管理 (Prompt Management)_**](https://arize.com/docs/phoenix/prompt-engineering/overview-prompts/prompt-management) - 通过版本控制、标签体系与科学实验，系统化管理并验证提示词变更。 - [**_PXI (Phoenix Intelligence)_**](https://arize.com/docs/phoenix/pxi) - 内置于 Phoenix 的 AI 研发智能体，辅助分析调用链路、迭代优化提示词及系统导航。 - [**_远程 MCP 服务器 (Remote MCP Server)_**](https://arize.com/docs/phoenix/integrations/remote-mcp) - 支持将 Claude Code、Cursor 及其他 MCP 客户端直连到 Phoenix 实例的 `/mcp` 端点，以便快捷检索链路追踪、数据集与实验指标。

Phoenix 与模型供应商和编程语言完全解耦，开箱即用地支持主流框架（包括 [OpenAI Agents SDK](https://arize.com/docs/phoenix/tracing/integrations-tracing/openai-agents-sdk)、[Claude Agent SDK](https://arize.com/docs/phoenix/integrations/python/claude-agent-sdk)、[LangGraph](https://arize.com/docs/phoenix/tracing/integrations-tracing/langchain)、[Vercel AI SDK](https://arize.com/docs/phoenix/tracing/integrations-tracing/vercel-ai-sdk)、[Mastra](https://arize.com/docs/phoenix/integrations/typescript/mastra)、[CrewAI](https://arize.com/docs/phoenix/tracing/integrations-tracing/crewai)、[LlamaIndex](https://arize.com/docs/phoenix/tracing/integrations-tracing/llamaindex)、[DSPy](https://arize.com/docs/phoenix/tracing/integrations-tracing/dspy)）以及大模型供应商（包括 [OpenAI](https://arize.com/docs/phoenix/tracing/integrations-tracing/openai)、[Anthropic](https://arize.com/docs/phoenix/tracing/integrations-tracing/anthropic)、[Google GenAI](https://arize.com/docs/phoenix/tracing/integrations-tracing/google-gen-ai)、[Google ADK](https://arize.com/docs/phoenix/integrations/llm-providers/google-gen-ai/google-adk-tracing)、[AWS Bedrock](https://arize.com/docs/phoenix/tracing/integrations-tracing/bedrock)、[OpenRouter](https://arize.com/docs/phoenix/integrations/python/openrouter)、[LiteLLM](https://arize.com/docs/phoenix/tracing/integrations-tracing/litellm) 等）。关于自动插桩的详细实现，请参阅 [OpenInference](https://github.com/Arize-ai/openinference) 项目。 Phoenix 可在几乎任何环境中运行，包括本地开发机、容器化环境或云端平台。详细操作请参阅[运行环境指南 (Environments)](https://arize.com/docs/phoenix/environments)，或直接查阅[链路追踪快速入门 (Tracing Quickstart)](https://arize.com/docs/phoenix/get-started/get-started-tracing)。 > [!NOTE] > Phoenix 为开源且支持私有化自托管。对于全托管生产级平台需求，Arize 提供了 [Arize AX](https://arize.com/products/ax/)——二者使用相同的 OpenTelemetry / OpenInference 插桩体系。详见[完整功能对比说明](https://arize.com/docs/phoenix/resources/frequently-asked-questions/what-is-the-difference-between-phoenix-and-arize)。 ## 目录 - [本地运行](#本地运行) - [为应用配置链路追踪](#为应用配置链路追踪) - [部署指南](#部署指南) - [软件包矩阵](#软件包矩阵) - [追踪集成生态](#追踪集成生态) - [安全沙箱环境](#安全沙箱环境-sandboxes) - [面向开发者与 AI 编程智能体](#面向开发者与-ai-编程智能体) - [安全性与隐私保护](#安全性与隐私保护) - [社区与交流](#社区与交流) ## 本地运行 通过 `pip` 或 `conda` 即可安装功能完整的 Phoenix。完整的安装与托管选项请参阅[安装指南](https://arize.com/docs/phoenix)。 ```shell pip install arize-phoenix phoenix serve ``` 或者使用 [`uvx`](https://docs.astral.sh/uv/) 实现免安装直接运行： ```shell uvx arize-phoenix serve ``` ## 为应用配置链路追踪 发送追踪数据最便捷的方式是让你的 [AI 编程智能体](https://arize.com/docs/phoenix/integrations/developer-tools/coding-agents)（如 Claude Code、Codex、Cursor 等）自动为你的应用程序添加插桩。在你的项目目录下直接运行： ```shell npx @arizeai/phoenix-cli setup # 或者在已安装 Phoenix 的环境下运行：px setup ``` Setup 命令会自动检测你使用的框架和 LLM 提供商，安装适用的 [OpenInference](https://github.com/Arize-ai/openinference) 插桩依赖，并配置好 Trace 导出通道。如果你更倾向于在代码中手动配置，请参阅[链路追踪技术文档](https://arize.com/docs/phoenix/tracing/llm-traces)。 ## 部署指南 Phoenix 容器镜像已发布在 [Docker Hub](https://hub.docker.com/r/arizephoenix/phoenix)，支持通过 Docker 部署，或通过 [Helm chart](https://github.com/Arize-ai/phoenix/tree/main/helm) 部署到 Kubernetes 集群。 关于 Docker Compose、Kubernetes / Helm 以及其他部署方案的完整介绍，请参阅[私有化自托管文档 (Self-Hosting)](https://arize.com/docs/phoenix/self-hosting)。

Deploy on Railway   Deploy to Render   Run on Google Cloud   Deploy to Azure   Deploy to AWS

> [!NOTE] > Google Cloud 按钮会在 Cloud Shell 中从源码构建 Phoenix，而非直接拉取预构建的 Docker Hub 镜像。Azure 部署模板提供的是纯 HTTP 服务（Azure Container Instances 默认不终止 TLS）——在生产使用前请在其前端配置 TLS 反向代理（例如 Application Gateway）。 ## 软件包矩阵 `arize-phoenix` 主包包含了完整的 Phoenix 平台。此外，若你已经独立部署了 Phoenix 服务端，我们还提供了轻量级的 Python 子包与 TypeScript 客户端包，以便与服务端协同使用。 ### Python 独立子包 | 软件包 | 版本与文档 | 描述 | | --------------------------------------------------------------------------------------------- | --------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- | ------------------------------------------------------------------------------------------ | | [arize-phoenix-otel](https://github.com/Arize-ai/phoenix/tree/main/packages/phoenix-otel) | [![PyPI Version](https://img.shields.io/pypi/v/arize-phoenix-otel)](https://pypi.org/project/arize-phoenix-otel/) [![Docs](https://img.shields.io/badge/docs-blue?logo=readthedocs&logoColor=white)](https://arize-phoenix.readthedocs.io/projects/otel/en/latest/index.html) | 对 OpenTelemetry 原语的轻量级封装，预置适配 Phoenix 的默认配置 | | [arize-phoenix-client](https://github.com/Arize-ai/phoenix/tree/main/packages/phoenix-client) | [![PyPI Version](https://img.shields.io/pypi/v/arize-phoenix-client)](https://pypi.org/project/arize-phoenix-client/) [![Docs](https://img.shields.io/badge/docs-blue?logo=readthedocs&logoColor=white)](https://arize-phoenix.readthedocs.io/projects/client/en/latest/index.html) | 通过 OpenAPI REST 接口与 Phoenix 服务器交互的轻量级客户端 | | [arize-phoenix-evals](https://github.com/Arize-ai/phoenix/tree/main/packages/phoenix-evals) | [![PyPI Version](https://img.shields.io/pypi/v/arize-phoenix-evals)](https://pypi.org/project/arize-phoenix-evals/) [![Docs](https://img.shields.io/badge/docs-blue?logo=readthedocs&logoColor=white)](https://arize-phoenix.readthedocs.io/projects/evals/en/latest/index.html) | 用于评估 LLM 应用程序的工具集，涵盖 RAG 相关性、回答精准度等维度评估 | ### TypeScript 独立子包 | 软件包 | 版本与文档 | 描述 | | --------------------------------------------------------------------------------------------------- | -------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- | ------------------------------------------------------------------------------------------------------------ | | [@arizeai/phoenix-otel](https://github.com/Arize-ai/phoenix/tree/main/js/packages/phoenix-otel) | [![NPM Version](https://img.shields.io/npm/v/%40arizeai%2Fphoenix-otel)](https://www.npmjs.com/package/@arizeai/phoenix-otel) [![Docs](https://img.shields.io/badge/docs-blue?logo=typescript&logoColor=white)](https://arize-ai.github.io/phoenix/) | 对 OpenTelemetry 原语的轻量级封装，预置适配 Phoenix 的默认配置 | | [@arizeai/phoenix-client](https://github.com/Arize-ai/phoenix/tree/main/js/packages/phoenix-client) | [![NPM Version](https://img.shields.io/npm/v/%40arizeai%2Fphoenix-client)](https://www.npmjs.com/package/@arizeai/phoenix-client) [![Docs](https://img.shields.io/badge/docs-blue?logo=typescript&logoColor=white)](https://arize-ai.github.io/phoenix/) | 用于对接 Arize Phoenix API 的 TypeScript 客户端 | | [@arizeai/phoenix-evals](https://github.com/Arize-ai/phoenix/tree/main/js/packages/phoenix-evals) | [![NPM Version](https://img.shields.io/npm/v/%40arizeai%2Fphoenix-evals)](https://www.npmjs.com/package/@arizeai/phoenix-evals) [![Docs](https://img.shields.io/badge/docs-blue?logo=typescript&logoColor=white)](https://arize-ai.github.io/phoenix/) | 专为 LLM 应用打造的 TypeScript 评估库（Alpha 预览版） | | [@arizeai/phoenix-mcp](https://github.com/Arize-ai/phoenix/tree/main/js/packages/phoenix-mcp) | [![NPM Version](https://img.shields.io/npm/v/%40arizeai%2Fphoenix-mcp)](https://www.npmjs.com/package/@arizeai/phoenix-mcp) [![Docs](https://img.shields.io/badge/docs-blue?logo=markdown&logoColor=white)](./js/packages/phoenix-mcp/README.md) | 适用于老版本 Phoenix 的独立 stdio MCP 服务器（维护模式——已被 Phoenix 内置的[远程 MCP 服务器](https://arize.com/docs/phoenix/integrations/remote-mcp)所取代） | | [@arizeai/phoenix-cli](https://github.com/Arize-ai/phoenix/tree/main/js/packages/phoenix-cli) | [![NPM Version](https://img.shields.io/npm/v/%40arizeai%2Fphoenix-cli)](https://www.npmjs.com/package/@arizeai/phoenix-cli) [![Docs](https://img.shields.io/badge/docs-blue?logo=typescript&logoColor=white)](https://arize.com/docs/phoenix/sdk-api-reference/typescript/arizeai-phoenix-cli) | 用于获取 Traces、数据集和实验数据的 CLI 工具，供 Claude Code、Cursor 等编程智能体使用 | ## 追踪集成生态 Phoenix 基于 OpenTelemetry 深度打造，具备完全的供应商、语言与框架解耦能力。关于追踪插桩库及应用示例的详情，请参阅 [OpenInference](https://github.com/Arize-ai/openinference) 开源项目以及[官方集成文档](https://arize.com/docs/phoenix/integrations)。 **Python 集成** | | 目标框架/服务 | 插桩软件包 | 版本 | |:---:|---|---|---| | | [OpenAI](https://arize.com/docs/phoenix/tracing/integrations-tracing/openai) | `openinference-instrumentation-openai` | [![PyPI Version](https://img.shields.io/pypi/v/openinference-instrumentation-openai.svg)](https://pypi.python.org/pypi/openinference-instrumentation-openai) | | | [OpenAI Agents](https://arize.com/docs/phoenix/tracing/integrations-tracing/openai-agents-sdk) | `openinference-instrumentation-openai-agents` | [![PyPI Version](https://img.shields.io/pypi/v/openinference-instrumentation-openai-agents.svg)](https://pypi.python.org/pypi/openinference-instrumentation-openai-agents) | | | [LlamaIndex](https://arize.com/docs/phoenix/tracing/integrations-tracing/llamaindex) | `openinference-instrumentation-llama-index` | [![PyPI Version](https://img.shields.io/pypi/v/openinference-instrumentation-llama-index.svg)](https://pypi.python.org/pypi/openinference-instrumentation-llama-index) | | | [DSPy](https://arize.com/docs/phoenix/tracing/integrations-tracing/dspy) | `openinference-instrumentation-dspy` | [![PyPI Version](https://img.shields.io/pypi/v/openinference-instrumentation-dspy.svg)](https://pypi.python.org/pypi/openinference-instrumentation-dspy) | | | [AWS Bedrock](https://arize.com/docs/phoenix/tracing/integrations-tracing/bedrock) | `openinference-instrumentation-bedrock` | [![PyPI Version](https://img.shields.io/pypi/v/openinference-instrumentation-bedrock.svg)](https://pypi.python.org/pypi/openinference-instrumentation-bedrock) | | | [LangChain](https://arize.com/docs/phoenix/tracing/integrations-tracing/langchain) | `openinference-instrumentation-langchain` | [![PyPI Version](https://img.shields.io/pypi/v/openinference-instrumentation-langchain.svg)](https://pypi.python.org/pypi/openinference-instrumentation-langchain) | | | [LangGraph](https://arize.com/docs/phoenix/integrations/python/langgraph) | `openinference-instrumentation-langchain` | [![PyPI Version](https://img.shields.io/pypi/v/openinference-instrumentation-langchain.svg)](https://pypi.python.org/pypi/openinference-instrumentation-langchain) | | | [MistralAI](https://arize.com/docs/phoenix/tracing/integrations-tracing/mistralai) | `openinference-instrumentation-mistralai` | [![PyPI Version](https://img.shields.io/pypi/v/openinference-instrumentation-mistralai.svg)](https://pypi.python.org/pypi/openinference-instrumentation-mistralai) | | | [Cohere](https://arize.com/docs/phoenix/integrations/llm-providers/cohere) | `openinference-instrumentation-cohere` | [![PyPI Version](https://img.shields.io/pypi/v/openinference-instrumentation-cohere.svg)](https://pypi.python.org/pypi/openinference-instrumentation-cohere) | | | [Together AI](https://arize.com/docs/phoenix/integrations/llm-providers/together) | `openinference-instrumentation-together` | [![PyPI Version](https://img.shields.io/pypi/v/openinference-instrumentation-together.svg)](https://pypi.python.org/pypi/openinference-instrumentation-together) | | | [Ollama](https://arize.com/docs/phoenix/integrations/llm-providers/ollama) | `openinference-instrumentation-ollama` | [![PyPI Version](https://img.shields.io/pypi/v/openinference-instrumentation-ollama.svg)](https://pypi.python.org/pypi/openinference-instrumentation-ollama) | | | [Google GenAI](https://arize.com/docs/phoenix/tracing/integrations-tracing/google-gen-ai) | `openinference-instrumentation-google-genai` | [![PyPI Version](https://img.shields.io/pypi/v/openinference-instrumentation-google-genai.svg)](https://pypi.python.org/pypi/openinference-instrumentation-google-genai) | | | [Google ADK](https://arize.com/docs/phoenix/integrations/llm-providers/google-gen-ai/google-adk-tracing) | `openinference-instrumentation-google-adk` | [![PyPI Version](https://img.shields.io/pypi/v/openinference-instrumentation-google-adk.svg)](https://pypi.python.org/pypi/openinference-instrumentation-google-adk) | | | [Guardrails](https://arize.com/docs/phoenix/tracing/integrations-tracing/guardrails) | `openinference-instrumentation-guardrails` | [![PyPI Version](https://img.shields.io/pypi/v/openinference-instrumentation-guardrails.svg)](https://pypi.python.org/pypi/openinference-instrumentation-guardrails) | | | [VertexAI](https://arize.com/docs/phoenix/tracing/integrations-tracing/vertexai) | `openinference-instrumentation-vertexai` | [![PyPI Version](https://img.shields.io/pypi/v/openinference-instrumentation-vertexai.svg)](https://pypi.python.org/pypi/openinference-instrumentation-vertexai) | | | [CrewAI](https://arize.com/docs/phoenix/tracing/integrations-tracing/crewai) | `openinference-instrumentation-crewai` | [![PyPI Version](https://img.shields.io/pypi/v/openinference-instrumentation-crewai.svg)](https://pypi.python.org/pypi/openinference-instrumentation-crewai) | | | [Haystack](https://arize.com/docs/phoenix/tracing/integrations-tracing/haystack) | `openinference-instrumentation-haystack` | [![PyPI Version](https://img.shields.io/pypi/v/openinference-instrumentation-haystack.svg)](https://pypi.python.org/pypi/openinference-instrumentation-haystack) | | | [LiteLLM](https://arize.com/docs/phoenix/tracing/integrations-tracing/litellm) | `openinference-instrumentation-litellm` | [![PyPI Version](https://img.shields.io/pypi/v/openinference-instrumentation-litellm.svg)](https://pypi.python.org/pypi/openinference-instrumentation-litellm) | | | [OpenRouter](https://arize.com/docs/phoenix/integrations/llm-providers/openrouter) | `openinference-instrumentation-openai` | [![PyPI Version](https://img.shields.io/pypi/v/openinference-instrumentation-openai.svg)](https://pypi.python.org/pypi/openinference-instrumentation-openai) | | | [OrcaRouter](https://arize.com/docs/phoenix/integrations/llm-providers/orcarouter) | `openinference-instrumentation-openai` | [![PyPI Version](https://img.shields.io/pypi/v/openinference-instrumentation-openai.svg)](https://pypi.python.org/pypi/openinference-instrumentation-openai) | | | [Groq](https://arize.com/docs/phoenix/tracing/integrations-tracing/groq) | `openinference-instrumentation-groq` | [![PyPI Version](https://img.shields.io/pypi/v/openinference-instrumentation-groq.svg)](https://pypi.python.org/pypi/openinference-instrumentation-groq) | | | [Instructor](https://arize.com/docs/phoenix/tracing/integrations-tracing/instructor) | `openinference-instrumentation-instructor` | [![PyPI Version](https://img.shields.io/pypi/v/openinference-instrumentation-instructor.svg)](https://pypi.python.org/pypi/openinference-instrumentation-instructor) | | | [Anthropic](https://arize.com/docs/phoenix/tracing/integrations-tracing/anthropic) | `openinference-instrumentation-anthropic` | [![PyPI Version](https://img.shields.io/pypi/v/openinference-instrumentation-anthropic.svg)](https://pypi.python.org/pypi/openinference-instrumentation-anthropic) | | | [Smolagents](https://huggingface.co/docs/smolagents/en/tutorials/inspect_runs) | `openinference-instrumentation-smolagents` | [![PyPI Version](https://img.shields.io/pypi/v/openinference-instrumentation-smolagents.svg)](https://pypi.python.org/pypi/openinference-instrumentation-smolagents) | | | [Agno](https://arize.com/docs/phoenix/tracing/integrations-tracing/agno) | `openinference-instrumentation-agno` | [![PyPI Version](https://img.shields.io/pypi/v/openinference-instrumentation-agno.svg)](https://pypi.python.org/pypi/openinference-instrumentation-agno) | | | [BeeAI](https://arize.com/docs/phoenix/integrations/python/beeai) | `openinference-instrumentation-beeai` | [![PyPI Version](https://img.shields.io/pypi/v/openinference-instrumentation-beeai.svg)](https://pypi.python.org/pypi/openinference-instrumentation-beeai) | | | [Strands Agents](https://arize.com/docs/phoenix/integrations/python/strands-agents) | `openinference-instrumentation-strands-agents` | [![PyPI Version](https://img.shields.io/pypi/v/openinference-instrumentation-strands-agents.svg)](https://pypi.python.org/pypi/openinference-instrumentation-strands-agents) | | | [Restate](https://arize.com/docs/phoenix/integrations/python/restate) | `openinference-instrumentation-openai-agents` | [![PyPI Version](https://img.shields.io/pypi/v/openinference-instrumentation-openai-agents.svg)](https://pypi.python.org/pypi/openinference-instrumentation-openai-agents) | | | [MCP](https://arize.com/docs/phoenix/tracing/integrations-tracing/model-context-protocol-mcp) | `openinference-instrumentation-mcp` | [![PyPI Version](https://img.shields.io/pypi/v/openinference-instrumentation-mcp.svg)](https://pypi.python.org/pypi/openinference-instrumentation-mcp) | | | [Pydantic AI](https://arize.com/docs/phoenix/integrations/python/pydantic) | `openinference-instrumentation-pydantic-ai` | [![PyPI Version](https://img.shields.io/pypi/v/openinference-instrumentation-pydantic-ai.svg)](https://pypi.python.org/pypi/openinference-instrumentation-pydantic-ai) | | | [AG2](https://arize.com/docs/phoenix/integrations/python/ag2) | `openinference-instrumentation-ag2` | [![PyPI Version](https://img.shields.io/pypi/v/openinference-instrumentation-ag2.svg)](https://pypi.python.org/pypi/openinference-instrumentation-ag2) | | | [Autogen AgentChat](https://arize.com/docs/phoenix/integrations/frameworks/autogen/autogen-tracing) | `openinference-instrumentation-autogen-agentchat` | [![PyPI Version](https://img.shields.io/pypi/v/openinference-instrumentation-autogen-agentchat.svg)](https://pypi.python.org/pypi/openinference-instrumentation-autogen-agentchat) | | | [Portkey](https://arize.com/docs/phoenix/integrations/portkey) | `openinference-instrumentation-portkey` | [![PyPI Version](https://img.shields.io/pypi/v/openinference-instrumentation-portkey.svg)](https://pypi.python.org/pypi/openinference-instrumentation-portkey) | | | [Agent Spec](https://arize.com/docs/phoenix/tracing/integrations-tracing/agentspec) | `openinference-instrumentation-agentspec` | [![PyPI Version](https://img.shields.io/pypi/v/openinference-instrumentation-agentspec.svg)](https://pypi.python.org/pypi/openinference-instrumentation-agentspec) | | | [Claude Agent SDK](https://arize.com/docs/phoenix/integrations/python/claude-agent-sdk) | `openinference-instrumentation-claude-agent-sdk` | [![PyPI Version](https://img.shields.io/pypi/v/openinference-instrumentation-claude-agent-sdk.svg)](https://pypi.python.org/pypi/openinference-instrumentation-claude-agent-sdk) | ## 跨度处理器 (Span Processors) 通过接入统一数据的 Span 处理器，标准化并转换来自其他第三方插桩库的观测数据： | 软件包 | 描述 | 版本 | | ------------------------------------------------------------------------------------------------- | ---------------------------------------------------------------- | ---------------------------------------------------------------------------------------------------------------------------------------------------------------------- | | [`openinference-instrumentation-openlit`](https://pypi.org/project/openinference-instrumentation-openlit/) | 面向 OpenLIT 追踪数据的 OpenInference Span 处理器。 | [![PyPI Version](https://img.shields.io/pypi/v/openinference-instrumentation-openlit.svg)](https://pypi.python.org/pypi/openinference-instrumentation-openlit) | | [`openinference-instrumentation-openllmetry`](https://pypi.org/project/openinference-instrumentation-openllmetry/) | 面向 OpenLLMetry (Traceloop) 追踪数据的 OpenInference Span 处理器。 | [![PyPI Version](https://img.shields.io/pypi/v/openinference-instrumentation-openllmetry.svg)](https://pypi.python.org/pypi/openinference-instrumentation-openllmetry) | ### JavaScript 集成 | | 目标框架/SDK | 软件包 | 版本 | | :-------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------: | ------------------------------------------------------------------------------------------- | --------------------------------------------------------- | --------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- | | | [OpenAI](https://arize.com/docs/phoenix/tracing/integrations-tracing/openai-node-sdk) | `@arizeai/openinference-instrumentation-openai` | [![NPM Version](https://img.shields.io/npm/v/@arizeai/openinference-instrumentation-openai.svg)](https://www.npmjs.com/package/@arizeai/openinference-instrumentation-openai) | | | [OpenAI Agents](https://arize.com/docs/phoenix/integrations/typescript/openai-agents) | `@arizeai/openinference-instrumentation-openai-agents` | [![NPM Version](https://img.shields.io/npm/v/@arizeai%2Fopeninference-instrumentation-openai-agents.svg)](https://www.npmjs.com/package/@arizeai/openinference-instrumentation-openai-agents) | | | [LangChain.js](https://arize.com/docs/phoenix/tracing/integrations-tracing/langchain) | `@arizeai/openinference-instrumentation-langchain` | [![NPM Version](https://img.shields.io/npm/v/@arizeai/openinference-instrumentation-langchain.svg)](https://www.npmjs.com/package/@arizeai/openinference-instrumentation-langchain) | | | [TanStack AI](https://arize.com/docs/phoenix/integrations/typescript/tanstack-ai) | `@arizeai/openinference-tanstack-ai` | [![NPM Version](https://img.shields.io/npm/v/@arizeai/openinference-tanstack-ai.svg)](https://www.npmjs.com/package/@arizeai/openinference-tanstack-ai) | | | [Vercel AI SDK](https://arize.com/docs/phoenix/tracing/integrations-tracing/vercel-ai-sdk) | `@arizeai/openinference-vercel` | [![NPM Version](https://img.shields.io/npm/v/@arizeai/openinference-vercel)](https://www.npmjs.com/package/@arizeai/openinference-vercel) | | | [BeeAI](https://arize.com/docs/phoenix/tracing/integrations-tracing/beeai) | `@arizeai/openinference-instrumentation-beeai` | [![NPM Version](https://img.shields.io/npm/v/@arizeai/openinference-instrumentation-beeai.svg)](https://www.npmjs.com/package/@arizeai/openinference-instrumentation-beeai) | | | [Claude Agent SDK](https://arize.com/docs/phoenix/integrations/typescript/claude-agent-sdk) | `@arizeai/openinference-instrumentation-claude-agent-sdk` | [![NPM Version](https://img.shields.io/npm/v/@arizeai%2Fopeninference-instrumentation-claude-agent-sdk.svg)](https://www.npmjs.com/package/@arizeai/openinference-instrumentation-claude-agent-sdk) | | | [Mastra](https://arize.com/docs/phoenix/integrations/typescript/mastra) | `@mastra/arize` | [![NPM Version](https://img.shields.io/npm/v/@mastra/arize.svg)](https://www.npmjs.com/package/@mastra/arize) | | | [MCP](https://arize.com/docs/phoenix/integrations/typescript/mcp) | `@arizeai/openinference-instrumentation-mcp` | [![NPM Version](https://img.shields.io/npm/v/@arizeai%2Fopeninference-instrumentation-mcp.svg)](https://www.npmjs.com/package/@arizeai/openinference-instrumentation-mcp) | ### Java 集成 | | 目标框架/生态 | 软件包 | 版本 | | :-------------------------------------------------------------------------------------------------: | ------------------------------------------------------------------------------------------------- | ------------------------------------------------------- | ----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- | | | [LangChain4j](https://github.com/Arize-ai/openinference/tree/main/java/instrumentation/openinference-instrumentation-langchain4j) | `openinference-instrumentation-langchain4j` | [![Maven Central](https://img.shields.io/maven-central/v/com.arize/openinference-instrumentation-langchain4j.svg)](https://central.sonatype.com/artifact/com.arize/openinference-instrumentation-langchain4j) | | | SpringAI | `openinference-instrumentation-springAI` | [![Maven Central](https://img.shields.io/maven-central/v/com.arize/openinference-instrumentation-springAI.svg)](https://central.sonatype.com/artifact/com.arize/openinference-instrumentation-springAI) | | | Spring AI 适用的 [Arconia](https://docs.arconia.io/arconia/latest/observability/semantic-conventions/openinference/) | `io.arconia:arconia-openinference-semantic-conventions` | [![Maven Central](https://img.shields.io/maven-central/v/io.arconia/arconia-openinference-ai-semantic-conventions.svg)](https://central.sonatype.com/artifact/io.arconia/arconia-openinference-ai-semantic-conventions) | ### Go 集成 | | 目标 SDK | 软件包 | 版本 | | :-------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------: | ------------------------------------------------------------------------------------------------------------------ | ------------------------------------------------------------------------------------- | ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- | | | [OpenAI](https://github.com/Arize-ai/openinference/tree/main/go/openinference-instrumentation-openai-go) | `github.com/Arize-ai/openinference/go/openinference-instrumentation-openai-go` | [![Go Reference](https://pkg.go.dev/badge/github.com/Arize-ai/openinference/go/openinference-instrumentation-openai-go.svg)](https://pkg.go.dev/github.com/Arize-ai/openinference/go/openinference-instrumentation-openai-go) | | | [Anthropic](https://github.com/Arize-ai/openinference/tree/main/go/openinference-instrumentation-anthropic-sdk-go) | `github.com/Arize-ai/openinference/go/openinference-instrumentation-anthropic-sdk-go` | [![Go Reference](https://pkg.go.dev/badge/github.com/Arize-ai/openinference/go/openinference-instrumentation-anthropic-sdk-go.svg)](https://pkg.go.dev/github.com/Arize-ai/openinference/go/openinference-instrumentation-anthropic-sdk-go) | ### 平台级集成 (Platforms) | | 平台名称 | 平台描述 | 文档指南 | | :----------------------------------------------------------------------------------------------: | -------------------------------------------------------------------------------------------------------- | -------------------------------------------------------------- | ----------------------------------------------------------------------------------------------------------------- | | | [BeeAI](https://docs.beeai.dev/observability/agents-traceability) | 内置可观测性支持的 AI Agent 开发框架 | [集成指南](https://docs.beeai.dev/observability/agents-traceability) | | | [Dify](https://docs.dify.ai/en/guides/monitoring/integrate-external-ops-tools/integrate-phoenix) | 开源大语言模型应用开发与运维平台 | [集成指南](https://docs.dify.ai/en/guides/monitoring/integrate-external-ops-tools/integrate-phoenix) | | | [Envoy AI Gateway](https://github.com/envoyproxy/ai-gateway) | 基于 Envoy Proxy 构建的面向 AI 工作负载的网关 | [集成指南](https://github.com/envoyproxy/ai-gateway/tree/main/cmd/aigw#opentelemetry-setup-with-phoenix) | | | [LangFlow](https://arize.com/docs/phoenix/tracing/integrations-tracing/langflow) | 用于构建多智能体与 RAG 应用的可视化低代码框架 | [集成指南](https://arize.com/docs/phoenix/tracing/integrations-tracing/langflow) | | | [LiteLLM Proxy](https://docs.litellm.ai/docs/observability/phoenix_integration#using-with-litellm-proxy) | 统一的 LLM 负载均衡与路由代理服务器 | [集成指南](https://docs.litellm.ai/docs/observability/phoenix_integration#using-with-litellm-proxy) | | | [Flowise](https://arize.com/docs/phoenix/integrations/platforms/flowise) | 用于快速搭建 LLM 应用的可视化拖拽编排平台 | [集成指南](https://arize.com/docs/phoenix/integrations/platforms/flowise) | | | [Prompt Flow](https://arize.com/docs/phoenix/integrations/platforms/prompt-flow) | 微软推出的 Prompt 工程与工作流编排套件 | [集成指南](https://arize.com/docs/phoenix/integrations/platforms/prompt-flow) | | | [NVIDIA NeMo](https://arize.com/docs/phoenix/integrations/python/nvidia) | 面向企业级智能体的 NVIDIA NeMo Agent Toolkit | [集成指南](https://arize.com/docs/phoenix/integrations/python/nvidia) | | | [Graphite](https://arize.com/docs/phoenix/integrations/python/graphite) | 配备可视化构建器的多智能体 LLM 工作流框架 | [集成指南](https://arize.com/docs/phoenix/integrations/python/graphite) | ## 安全沙箱环境 (Sandboxes) 在云端托管沙箱服务中运行 Phoenix [代码评估器 (Code Evaluators)](https://arize.com/docs/phoenix/evaluation/server-evals/code-evaluators)，实现内核级安全隔离与运行时依赖动态安装。 | 集成提供商 | 平台描述 | 文档指引 | | --------------------------------------------------------------------------------- | ---------------------------------------------------------------- | ---------------------------------------------------------------------------------- | | [E2B](https://arize.com/docs/phoenix/integrations/sandboxes/e2b) | 面向 AI 生成代码的托管式 Micro-VM 微虚拟机安全沙箱 | [集成指南](https://arize.com/docs/phoenix/integrations/sandboxes/e2b) | | [Daytona](https://arize.com/docs/phoenix/integrations/sandboxes/daytona) | 支持快照秒级启动的托管开发沙箱环境 | [集成指南](https://arize.com/docs/phoenix/integrations/sandboxes/daytona) | | [Vercel Sandbox](https://arize.com/docs/phoenix/integrations/sandboxes/vercel) | 运行在 Vercel 底层基础设施上的按需瞬态计算沙箱 | [集成指南](https://arize.com/docs/phoenix/integrations/sandboxes/vercel) | | [Modal](https://arize.com/docs/phoenix/integrations/sandboxes/modal) | 无服务器（Serverless）、Python 原生的容器计算平台 | [集成指南](https://arize.com/docs/phoenix/integrations/sandboxes/modal) | ## 面向开发者与 AI 编程智能体 Phoenix 既为人类工程师打造，也原生支持各类 AI 编程智能体协作交互。以下三套核心界面使智能体（如 Claude Code、Codex、Cursor 等）能够深度感知并操作你的 Traces、数据集和实验： - **[CLI 命令行](https://arize.com/docs/phoenix/sdk-api-reference/typescript/arizeai-phoenix-cli)** — 运行 `npx @arizeai/phoenix-cli` 抓取 Traces、数据集与实验指标，并自动插桩应用（`setup`），使智能体可以直接在终端获取上下文并执行任务。 - **[智能体技能 (Skills)](https://docs.anthropic.com/en/docs/claude-code/skills)** — [`.agents/skills/`](.agents/skills/) 目录打包了专有工作流，赋予编程智能体使用 Phoenix 调试、评估与链路追踪的完整技能。 - **[远程 MCP 服务器](https://arize.com/docs/phoenix/integrations/remote-mcp)** — 支持将任意符合 MCP 规范的客户端连接至 Phoenix 实例的 `/mcp` 端点，直接执行上下文检索与操作。 关于智能体的完整配置与调用方式，请参阅[编程智能体集成指南](https://arize.com/docs/phoenix/integrations/developer-tools/coding-agents)。 | Agent Skill 技能库 | 技能功能描述 | | -------------------------------------------------- | -------------------------------------------------------------------------------------------------------------------------- | | [phoenix-cli](.agents/skills/phoenix-cli/) | 使用 Phoenix CLI 调试 LLM 应用——获取链路追踪、分析执行错误、审查对比实验并查询 GraphQL API | | [phoenix-evals](.agents/skills/phoenix-evals/) | 使用 Phoenix 构建并运行面向 AI / LLM 应用的多维度评估器 | | [phoenix-tracing](.agents/skills/phoenix-tracing/) | 面向大模型应用的 OpenInference 语义约定与全链路追踪插桩规范 | ## 安全性与隐私保护 我们高度重视用户的数据安全与隐私保护。详情请参阅我们的[安全性与隐私政策文档](https://arize.com/docs/phoenix/self-hosting/security/privacy)。 ### 遥测说明 默认情况下，Phoenix 仅收集基础的 Web 访问分析数据（例如页面浏览量、UI 交互事件），以协助我们理解产品使用情况并持续改进体验。**Phoenix 绝不会收集你的任何 Trace 调用数据、评估结果或敏感业务信息。** 如需完全关闭遥测，只需设置环境变量：`PHOENIX_TELEMETRY_ENABLED=false` ## 社区与交流 加入我们的开源社区，与数千名 AI 开发者共同交流探讨： - 🌍 加入官方 [Slack 开发者社区](https://join.slack.com/t/arize-ai/shared_invite/zt-3r07iavnk-ammtATWSlF0pSrd1DsMW7g)。 - 📚 查阅完整的 [官方在线技术文档](https://arize.com/docs/phoenix)。 - 💡 在 _#phoenix-support_ 频道提问与交流反馈。 - 🌟 在 [GitHub](https://github.com/Arize-ai/phoenix) 仓库为我们点亮 Star。 - 🐞 通过 [GitHub Issues](https://github.com/Arize-ai/phoenix/issues) 反馈缺陷与功能建议。 - 𝕏 在 [𝕏 (Twitter)](https://twitter.com/ArizePhoenix) 上关注最新动态。 - 💼 在 [LinkedIn](https://www.linkedin.com/showcase/113218220) 上关注官方账号。 - 🗺️ 查看 [公开路线图 (Roadmap)](https://github.com/orgs/Arize-ai/projects/45) 了解未来演进方向。 - 🧑‍🏫 在 Arize Learning Hub 深入学习 [AI 智能体](http://arize.com/ai-agents/) 与 [LLM 效果评估](https://arize.com/llm-evaluation)。 ## 重大变更 (Breaking Changes) 关于历史版本的重大架构调整与迁移指南，请参阅[迁移指南 (Migration Guide)](./MIGRATION.md)。 ## 版权、专利与许可证 Copyright 2025 Arize AI, Inc. All Rights Reserved. 本软件的部分代码受一项或多项美国专利保护。详情请参阅 [IP_NOTICE](https://github.com/Arize-ai/phoenix/blob/main/IP_NOTICE)。 本项目采用 Elastic License 2.0 (ELv2) 许可证开源。详情请参阅 [LICENSE](https://github.com/Arize-ai/phoenix/blob/main/LICENSE)。 --- > 💡 **文档维护说明**：本中文文档由社区志愿者（@JasonYeYuhe）翻译维护，最后同步更新于 2026年9月8日。如发现内容与官方英文原版存在差异或新特性滞后，欢迎提交 PR 共同完善！