LangSmith Monitor Skill

Monitor LangSmith traces and runs using the LangSmithMonitor class, which fetches recent runs, retrieves project statistics, and inspects trace details via API calls. Instantiate with your LANGCHAIN_API_KEY and call methods like get_recent_runs() or get_trace_details() to return structured JSON data for dashboards and diagnostics.

Yes. The LangSmith monitoring Skill enables cost awareness by retrieving project statistics that include token usage metrics. Call get_project_stats() to access cost and token data across your LangSmith projects.

Call the monitor_health() method on the LangSmithMonitor instance to perform health checks and verify system connectivity. The method returns standardized JSON responses indicating the status of your LangSmith connection.

Yes. The Skill supports monitoring LangSmith traces and runs across both LLM pipelines and reinforcement learning training workflows, enabling health checks and performance metric retrieval for both use cases.

Set up monitoring by configuring the LANGCHAIN_API_KEY environment variable. The LangSmithMonitor class uses this API key to authenticate with the LangSmith client and access traces, runs, and project metrics.

Retrieve detailed trace information by calling get_trace_details() on the LangSmithMonitor instance. This method returns structured trace data for inspection, helping identify issues in your LLM or RL pipeline execution.