langsmith-fetch

Fetch LangChain and LangGraph execution traces from LangSmith Studio for debugging.

4|1|Updated Mar 14, 2026

Search Tags:#debugging #langchain #agent #langgraph #traces #langsmith

Ready to run

npx skills add https://github.com/GPTtang/skill-atlas --skill langsmith-fetch-gpttang

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: langsmith-fetch
Source: https://github.com/GPTtang/skill-atlas/tree/main/skills/ai-agent/langsmith-fetch
Command: npx skills add https://github.com/GPTtang/skill-atlas --skill langsmith-fetch-gpttang

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires langsmith-fetch, and includes scripts (resource) and references (resource) components.

What problem does it solve?

This Skill helps developers debug LangChain and LangGraph agents by fetching and analyzing execution traces directly from LangSmith Studio, streamlining the process of identifying and resolving issues.

Core Features & Use Cases

Fetch Execution Traces: Retrieve recent or specific agent execution traces.
Analyze Errors & Performance: Identify failures, tool call issues, and performance bottlenecks.
Export Debug Sessions: Save traces and threads for detailed offline analysis or sharing.
Use Case: When an agent behaves unexpectedly, use this Skill to quickly pull the last 5 minutes of traces, identify any errors, and pinpoint the exact tool call that failed.

Quick Start

Debug my agent by showing me the last 5 minutes of traces in a pretty format.

Frequently Asked Questions about langsmith-fetch

High-intent search queries and answers about installing and using this skill.

FAQPage Schema

How do I debug LangChain agent errors from execution traces?▼

Debug LangChain agent errors by fetching execution traces from LangSmith Studio to identify failures, tool call issues, and memory operations. This process pinpoints the exact tool call that failed when an agent behaves unexpectedly.

How do I fetch recent LangSmith traces to analyze agent performance?▼

Fetch recent LangSmith traces by requesting the last 5 minutes of execution data to analyze agent performance. This retrieves specific traces to identify performance bottlenecks and resolve unexpected agent behavior.

Can I export LangGraph debug sessions for offline analysis?▼

You can export LangGraph debug sessions by saving traces and threads for detailed offline analysis or sharing. This allows developers to review agent execution data outside the LangSmith Studio environment.

Do I need a LangSmith API key to fetch agent traces?▼

You need a configured LangSmith API key and project to fetch agent traces. The Skill requires the langsmith-fetch CLI to retrieve execution data and debug LangChain and LangGraph agents effectively.

What is the best way to debug LangGraph agent tool call issues?▼

The best way to debug LangGraph agent tool call issues is fetching execution traces from LangSmith Studio to analyze behavior. This identifies exact tool call failures and memory operation errors during agent execution.