Skip to main content 홈 크리에이터 jeremylongshore tons-of-skills-marketplace langfuse-sdk-patterns
langfuse-sdk-patterns Langfuse SDK best practices, patterns, and idiomatic usage.
Use when learning Langfuse SDK patterns, implementing proper tracing,
or following best practices for LLM observability.
Trigger with phrases like "langfuse patterns", "langfuse best practices",
"langfuse SDK guide", "how to use langfuse", "langfuse idioms".
설치로 이동 Skills Marketplace 커뮤니티가 만든 AI 스킬을 발견하고 탐색하세요.
Codex 또는 Claude로 설치 이 Prompt를 복사해 Codex, Claude 또는 다른 어시스턴트에 붙여 넣으면 Skill 페이지를 검토하고 설치를 진행할 수 있습니다.
직접 명령은 검토 Prompt를 거치지 않습니다. 실행하기 전에 소스를 확인하세요.
npx skills add https://github.com/jeremylongshore/tons-of-skills-marketplace --skill langfuse-sdk-patterns명령은 한 줄로 유지됩니다. 복사하기 전에 가로로 스크롤해 전체 내용을 확인하세요.
로컬 사본을 원하시나요? SkillsMP에서 현재 제공할 수 있는 파일을 다운로드하세요.
Zip 다운로드 다운로드 중... 이 저장소의 다른 Skills langchain-deploy-integration Deploy a LangChain 1.0 / LangGraph 1.0 app to Cloud Run, Vercel, or LangServe correctly — with timeouts sized for chain length, cold-start mitigation, SSE anti-buffering headers, and Secret Manager over .env. Use when prepping a first production deploy, debugging a stream that hangs behind a proxy, or diagnosing p99 latency spikes. Trigger with "langchain deploy", "langchain cloud run", "langchain vercel python", "langchain langserve", or "langchain docker".
langchain-langgraph-agents Build a correct LangGraph 1.0 ReAct agent with create_react_agent — typed tools, error propagation, recursion caps, and stop conditions that actually stop. Use when writing a first tool-calling agent, migrating from AgentExecutor or initialize_agent, or diagnosing an agent that loops on vague prompts. Trigger with "langgraph agent", "create_react_agent", "langgraph tool calling", "AgentExecutor migration", or "agent loop cost".
langchain-langgraph-human-in-loop Build LangGraph 1.0 human-in-the-loop approval flows with interrupt_before /
interrupt_after and Command(resume=...) — JSON-serializable state, clean
resume semantics, and UI wiring for approval decisions. Use when adding an
approval gate before an expensive tool call, wiring a Slack/web UI for agent
approvals, or debugging a graph that crashes on interrupt.
Trigger with "langgraph human in loop", "langgraph interrupt_before",
"langgraph approval flow", "Command resume", "langgraph HITL".
jeremylongshore
jeremylongshore/tons-of-skills-marketplace
GitHub 저장소 열기 name langfuse-sdk-patterns description Langfuse SDK best practices, patterns, and idiomatic usage.
Use when learning Langfuse SDK patterns, implementing proper tracing,
or following best practices for LLM observability.
Trigger with phrases like "langfuse patterns", "langfuse best practices",
"langfuse SDK guide", "how to use langfuse", "langfuse idioms".
allowed-tools Read, Write, Edit version 1.12.0 license MIT author Jeremy Longshore <jeremy@intentsolutions.io> tags ["saas","langfuse","observability","llm","tracing"] compatibility Designed for Claude Code
Langfuse SDK Patterns
Overview
Production-quality patterns for the Langfuse SDK: singleton clients, the observe wrapper, startActiveObservation for nested traces, session tracking, graceful shutdown, and error-safe tracing.
Prerequisites
Completed langfuse-install-auth setup
Understanding of async/await patterns
For v4+: @langfuse/tracing, @langfuse/otel, @opentelemetry/sdk-node
Instructions
Pattern 1: Singleton Client with Graceful Shutdown
import { LangfuseClient } from "@langfuse/client" ;
import { LangfuseSpanProcessor } from "@langfuse/otel" ;
import { NodeSDK } from "@opentelemetry/sdk-node" ;
let client : LangfuseClient | null = null ;
export function getLangfuseClient (LangfuseClient {
if (!client) {
client = new LangfuseClient ();
}
return client;
}
let sdk : NodeSDK | null = null ;
export function initTracing (
NodeSDK
if
new
NodeSDK
spanProcessors
new
LangfuseSpanProcessor
start
const
shutdown
async
await
shutdown
exit
0
on
"SIGTERM"
on
"SIGINT"
return
import { Langfuse } from "langfuse" ;
let instance : Langfuse | null = null ;
export function getLangfuse (Langfuse {
if (!instance) {
instance = new Langfuse ({
flushAt : 15 ,
flushInterval : 10000 ,
});
process.on ("beforeExit" , () => instance?.shutdownAsync ());
}
return instance;
}
Pattern 2: observe Wrapper for Existing Functions The observe wrapper is the most ergonomic way to add tracing. It wraps any function and auto-creates a span.
import { observe, updateActiveObservation } from "@langfuse/tracing" ;
const fetchUserProfile = observe (async (userId : string ) => {
updateActiveObservation ({ input : { userId } });
const profile = await db.users .findById (userId);
updateActiveObservation ({ output : { found : !!profile } });
return profile;
});
const summarize = observe (
{ name : "summarize-text" , asType : "generation" },
async (text : string ) => {
updateActiveObservation ({ model : "gpt-4o-mini" , input : text });
const result = await openai.chat .completions .create ({
model : "gpt-4o-mini" ,
messages : [{ role : "user" , content : `Summarize: ${text} ` }],
});
const output = result.choices [0 ].message .content ;
updateActiveObservation ({
output,
usage : {
promptTokens : result.usage ?.prompt_tokens ,
completionTokens : result.usage ?.completion_tokens ,
},
});
return output;
}
);
const pipeline = observe (async (userId : string ) => {
const profile = await fetchUserProfile (userId);
const summary = await summarize (profile.bio );
return { profile, summary };
});
Pattern 3: startActiveObservation for Inline Control Use when you need fine-grained control over observation lifecycle within a function:
import { startActiveObservation, updateActiveObservation } from "@langfuse/tracing" ;
async function processOrder (orderId : string return await startActiveObservation ("process-order" , async () => {
updateActiveObservation ({ input : { orderId } });
const validated = await startActiveObservation ("validate" , async () => {
const result = await validateOrder (orderId);
updateActiveObservation ({ output : { valid : result.valid } });
return result;
});
if (!validated.valid ) {
updateActiveObservation ({ output : { error : "validation failed" } });
return { success : false };
}
const description = await startActiveObservation (
{ name : "generate-confirmation" , asType : "generation" },
async () => {
updateActiveObservation ({ model : "gpt-4o-mini" });
const result = await generateConfirmation (orderId);
updateActiveObservation ({ output : result });
return result;
}
);
updateActiveObservation ({ output : { success : true } });
return { success : true , description };
});
}
Pattern 4: Session and User Tracking Link traces across conversation turns for user-level analytics:
await startActiveObservation ("chat-turn" , async () => {
updateActiveObservation ({
metadata : {
sessionId : "session-abc-123" ,
userId : "user-456" ,
},
});
await handleUserMessage (message);
});
const trace = langfuse.trace ({
name : "chat-turn" ,
sessionId : "session-abc-123" ,
userId : "user-456" ,
input : { message },
});
Pattern 5: Error-Safe Tracing Never let tracing failures break your application:
import { observe, updateActiveObservation } from "@langfuse/tracing" ;
const safeObserve = <T extends (...args : any []) => Promise <any >>(
name : string ,
fn : T
): T =>return (async (...args : Parameters <T>) => {
try {
return await observe ({ name }, async () => {
updateActiveObservation ({ input : args });
const result = await fn (...args);
updateActiveObservation ({ output : result });
return result;
})();
} catch (tracingError) {
console .warn (`Tracing error in ${name} :` , tracingError);
return fn (...args);
}
}) as T;
};
const processRequest = safeObserve ("process-request" , async (input : string ) => {
return await callLLM (input);
});
Pattern 6: Legacy v3 -- Always End Spans
const span = trace.span ({ name : "risky-operation" , input : data });
try {
const result = await riskyOperation (data);
span.end ({ output : result });
return result;
} catch (error) {
span.end ({ level : "ERROR" , statusMessage : String (error) });
throw error;
}
Anti-Patterns to Avoid Anti-Pattern Problem Correct Pattern new Langfuse() per requestMemory leaks, duplicate traces Singleton client Awaiting flush in hot path Adds latency to every request Background flush, shutdown handler Logging full request bodies Trace payloads too large Truncate/summarize inputs Missing .end() on spans (v3) Spans show "in progress" forever Use try/finally or observe wrapper Hardcoding API keys Security risk Environment variables only
Resources
Next Steps For OpenAI/LangChain tracing examples, see langfuse-core-workflow-a.
