🔭
统一观测模型
一套 run / generation / tool span / score 抽象，与宿主系统解耦，多宿主接入无需重复造轮子。
Yingdao AI ObservabilityAgent 运行全链路可观测
把 Agent 的每一次 LLM 调用、工具执行和推理过程，映射为 Langfuse 可视化 trace
Skip to content
快速上手
bash
npm install @yingdao-ai/agent-observabilityts
import { AgentObserver, LangfuseDirectTransport } from "@yingdao-ai/agent-observability"
const observer = new AgentObserver({
transport: new LangfuseDirectTransport({
host: "https://langfuse.example.com",
publicKey: process.env.LANGFUSE_PUBLIC_KEY!,
secretKey: process.env.LANGFUSE_SECRET_KEY!,
environment: "prod",
}),
})
const run = observer.startRun({ id: "session-1", name: "my_agent", input: "用户输入" })
if (run) {
run.startGeneration({ model: "openai/gpt-4.1" })
run.appendText("assistant 输出")
run.endGeneration({ usage: { input: 100, output: 20, total: 120 } })
await observer.endRun("session-1", "最终输出")
await observer.gate.waitAll()
}