Debug your AI app with traces and test results.
Guides the setup of tracing, evaluation and monitoring so users can debug and test LLM applications with LangSmith.
When to use it
Use when debugging LLM applications, evaluating outputs, monitoring production systems or building regression tests for AI features.
Give it an LLM application to instrument; it adds code for LangSmith tracing, evaluation or monitoring.
What you provide
The SDK examples run in Python; no Python version is specified.
Uses the langsmith package, with version 0.2.0 or newer declared as a dependency; the quick start installs it.
This skill
LangSmith
Installs the LangSmith SDK
Uploads traces and manages datasets (irreversible)
OpenAI
Sends prompts to OpenAI