安装
使用您偏好的包管理器安装所需的包：pip install langsmith autogen-agentchat autogen-ext opentelemetry-instrumentation-openai
uv add langsmith autogen-agentchat autogen-ext opentelemetry-instrumentation-openai
设置
1. 配置环境变量
设置您的 API 密钥 和项目名称：export LANGSMITH_API_KEY=<your_langsmith_api_key>
export LANGSMITH_PROJECT=<your_project_name>
export OPENAI_API_KEY=<your_openai_api_key>
2. 配置 OpenTelemetry 集成
在您的 AutoGen 应用中，配置 LangSmith OpenTelemetry 集成以及 OpenAI 检测器：from langsmith.integrations.otel import OtelSpanProcessor
from opentelemetry import trace
from opentelemetry.sdk.trace import TracerProvider
from opentelemetry.instrumentation.openai import OpenAIInstrumentor
# 设置追踪提供者
tracer_provider = TracerProvider()
tracer_provider.add_span_processor(OtelSpanProcessor())
trace.set_tracer_provider(tracer_provider)
# 检测 OpenAI 调用
OpenAIInstrumentor().instrument()
3. 创建并运行您的 AutoGen 应用
配置完成后，您的 AutoGen 应用将自动向 LangSmith 发送追踪。将追踪提供者传递给运行时以获得完整的追踪覆盖：import asyncio
from autogen_agentchat.agents import AssistantAgent
from autogen_agentchat.conditions import MaxMessageTermination, TextMentionTermination
from autogen_agentchat.teams import SelectorGroupChat
from autogen_agentchat.ui import Console
from autogen_core import SingleThreadedAgentRuntime
from autogen_ext.models.openai import OpenAIChatCompletionClient
from langsmith.integrations.otel import OtelSpanProcessor
from opentelemetry import trace
from opentelemetry.sdk.trace import TracerProvider
from opentelemetry.instrumentation.openai import OpenAIInstrumentor
# 设置追踪
tracer_provider = TracerProvider()
tracer_provider.add_span_processor(OtelSpanProcessor())
trace.set_tracer_provider(tracer_provider)
OpenAIInstrumentor().instrument()
# 定义一个工具
def percentage_change(start: float, end: float) -> float:
"""计算两个值之间的百分比变化。"""
if start == 0:
return float("inf")
return ((end - start) / start) * 100
async def main():
model_client = OpenAIChatCompletionClient(model="gpt-4o")
tracer = trace.get_tracer("autogen-demo")
with tracer.start_as_current_span("run_team"):
planning_agent = AssistantAgent(
"PlanningAgent",
description="规划任务并委派。",
model_client=model_client,
system_message=(
"你是一个规划代理。规划并委派任务。\n"
"分配任务时，请使用：1. <agent> : <task>\n"
'任务完成后，进行总结并以 "TERMINATE" 结束。'
),
)
data_analyst = AssistantAgent(
"DataAnalystAgent",
description="执行计算。",
model_client=model_client,
tools=[percentage_change],
system_message="你是一个数据分析师。使用工具来计算结果。",
)
termination = TextMentionTermination("TERMINATE") | MaxMessageTermination(max_messages=25)
# 将 tracer_provider 传递给运行时
runtime = SingleThreadedAgentRuntime(tracer_provider=trace.get_tracer_provider())
runtime.start()
team = SelectorGroupChat(
[planning_agent, data_analyst],
model_client=model_client,
termination_condition=termination,
allow_repeated_speaker=True,
runtime=runtime,
)
task = "你开始时有 100 个苹果，现在你有 120 个苹果。百分比变化是多少？"
await Console(team.run_stream(task=task))
await runtime.stop()
await model_client.close()
if __name__ == "__main__":
asyncio.run(main())
高级用法
自定义元数据和标签
您可以通过设置跨度属性为追踪添加自定义元数据：from opentelemetry import trace
tracer = trace.get_tracer(__name__)
async def run_with_metadata():
with tracer.start_as_current_span("autogen_workflow") as span:
span.set_attribute("langsmith.metadata.session_type", "multi_agent")
span.set_attribute("langsmith.metadata.agent_count", "2")
span.set_attribute("langsmith.span.tags", "autogen,planning")
# 您的 AutoGen 代码在此处
await Console(team.run_stream(task=task))
与其他检测器结合使用
您可以将 AutoGen 追踪与其他 OpenTelemetry 检测器结合使用：from opentelemetry.instrumentation.openai import OpenAIInstrumentor
from opentelemetry.instrumentation.httpx import HTTPXClientInstrumentor
# 初始化多个检测器
OpenAIInstrumentor().instrument()
HTTPXClientInstrumentor().instrument()
资源
将这些文档 通过 MCP 连接到 Claude、VSCode 等，以获取实时答案。