Review Langfuse traces, observations, scores, and metrics to debug yuno.md issues.
npx skills add https://github.com/Xuno-io/yuno-md --skill langfuse-reviewer
Please help me install this Agent Skill. Skill: langfuse-reviewer Source: https://github.com/Xuno-io/yuno-md/tree/main/skill/langfuse-reviewer Command: npx skills add https://github.com/Xuno-io/yuno-md --skill langfuse-reviewer
This Skill helps developers and operators debug issues within the yuno.md application by providing direct access to Langfuse traces, observations, scores, and metrics.
curl command examples for common debugging scenarios like investigating user issues, finding errors, and analyzing costs/latency.Use the langfuse reviewer skill to find recent traces for the user with telegram ID 12345.
High-intent search queries and answers about installing and using this skill.
Debug Langfuse traces by reviewing observations, scores, and metrics to investigate user issues. Combine MCP tools with REST API calls to access traces, sessions, and metrics for comprehensive observability and pinpoint exact LLM calls causing errors.
LLM observability provides visibility into application behavior through trace review. Analyzing trace observations and metrics helps with error identification by exposing the exact LLM calls and prompt executions that led to unexpected application behavior.
Analyze latency and cost using Langfuse metrics by executing REST API calls to retrieve production trace data. Review specific observations and scores to isolate expensive or slow LLM calls within your application sessions.
You can review and manage Langfuse prompts during debugging workflows by integrating MCP tools. This allows you to access documentation, manage prompts, and review prompt executions alongside trace observations to resolve issues.
The best way to find a specific user session in Langfuse traces is by using REST API calls with user identifiers, such as a telegram ID. This retrieves associated sessions, traces, and observations to debug reported issues.
You need MCP tools to access Langfuse prompts and documentation, while REST API calls handle traces, observations, sessions, and metrics. Combining both provides comprehensive observability for debugging latency, cost, and error issues.
jurislmManage Langfuse prompts, traces, observations, and scores via MCP.
junchenghuoDefine SLO metrics, alert thresholds, and incident review processes.
aleonsaTrace LLM calls, spans, and costs in Python applications with Langfuse.
giosuetedeschi-specTrack LLM traces, spans, and metrics for production AI applications.