langsmith-fetch-skill

왜 쓰나요

핵심 기능

Pulls traces directly from LangSmith Studio via CLI
Auto-triggers on debugging-flavored prompts
Tracks failures + root causes
Latency + token usage analysis
Tool-call and memory operation review
Session export

라이브 데모

실제 사용 모습

langsmith-fetch-skill.replay ▶ 준비됨

0/0

설치

작성자 저장소 · Claude Skills 공식 문서

사용 사례

실전 활용법: langsmith-fetch-skill

Diagnose a slow agent without opening LangSmith UI

👤 LangChain/LangGraph devs ⏱ ~20 min intermediate

언제 쓸까: An agent that used to be snappy is now slow; you want root cause.

흐름

Ask Claude to diagnose

Why is my agent <name> slow? Pull the last 20 traces from LangSmith and find the hotspot.✓ 복사됨

→ Skill fetches traces; reports latency by step
Drill in

Focus on trace <id> — break down tool-call times.✓ 복사됨

→ Per-tool latency with the outlier called out

결과: A specific step to optimize, not a hunch.

함정

Optimizing the average when tail latency is the problem — Ask for p95/p99 specifically

함께 쓰기: filesystem

Root-cause an agent failure from a trace id

👤 Anyone running LangGraph in production ⏱ ~15 min intermediate

언제 쓸까: A user reported an error; you have the trace id.

흐름

Fetch + analyze

Debug trace <id>. What was the immediate error, what's the root cause in the graph?✓ 복사됨

→ Error step + upstream decision that led to it

결과: A fix targeting the actual decision point, not the symptom.

Find which node burns the most tokens

👤 Teams optimizing LLM spend ⏱ ~30 min intermediate

언제 쓸까: Monthly review of agent spend.

흐름

Aggregate token usage

Across last 7 days' traces, rank graph nodes by total token usage.✓ 복사됨

→ Ranked table with % share
Propose tightening

For the top 3, propose prompt-compression or model downgrade options.✓ 복사됨

→ Targeted opt list

결과: Measurable spend reduction without chasing tiny savings.

조합

다른 MCP와 조합해 10배 효율

langsmith-fetch-skill + filesystem

Save exports alongside the agent code for lineage

Export the debugging session to debugging/traces/<date>.json.✓ 복사됨

langsmith-fetch-skill + github

Attach trace summary to a bug report

Summarize trace <id> and post as comment on GitHub issue #42.✓ 복사됨

도구

이 MCP가 노출하는 것

도구	입력	언제 호출	비용
fetch_recent	project, limit	Broad diagnosis	LangSmith quota
fetch_trace	trace_id	Deep dive on one run	LangSmith quota
analyze_failures	traces	Post-fetch analysis	0
export_session	traces	Archive for later	0

비용 및 제한

운영 비용

API 쿼터: LangSmith: depends on plan; Studio has generous quotas
호출당 토큰: Low — traces are structured
금액: Free skill; LangSmith free tier covers most devs
팁: Use fetch_recent with a narrow window, not full history

보안

권한, 시크릿, 파급범위

최소 스코프: LangSmith: read on your project(s)

자격 증명 저장: LANGCHAIN_API_KEY in env

데이터 외부 송신: Only to LangSmith (where the traces already live)

절대 부여 금지: Write to LangSmith — skill only reads

Traces can include user inputs and outputs — treat them as user data under your privacy policy.

문제 해결

자주 발생하는 오류와 해결

langsmith-fetch: command not found

Install LangSmith CLI or ensure it's on PATH

확인: which langsmith-fetch

Empty results

Check project name and time window; traces are project-scoped

대안