skills.rest
yonatangross
yonatangrossCOMMUNITY
Explore all skills by this author →
𝕏@yonyonizGitHub

langfuse-observability

Trace LLM calls and track costs with Langfuse observability.

3|Updated Dec 27, 2025
Search Tags:#evaluation#tracing#langfuse#observe#cost-tracking#prompt-management#llm-observability
One-click install
npx skills add https://github.com/yonatangross/create-yg-app --skill langfuse-observability-yonatangross
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: langfuse-observability
Source: https://github.com/yonatangross/create-yg-app/tree/main/.claude/skills/langfuse-observability
Command: npx skills add https://github.com/yonatangross/create-yg-app --skill langfuse-observability-yonatangross

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

Langfuse observability provides end-to-end tracing, cost tracking, prompt management, and evaluation for LLM-powered applications, enabling production-grade visibility.

Core Features & Use Cases

  • Distributed tracing: automatic parent-child spans, multi-agent workflows
  • Cost tracking: token usage, per-model pricing, and dashboards
  • Prompt management: versioned prompts linked to traces for reproducibility
  • LLM evaluation: scores and dashboards for quality trends
  • Session & user analytics: group traces by session or user to track journeys

Quick Start

Instrument LLM calls with the @observe decorator, initialize Langfuse client, and explore traces in the UI.

Frequently Asked Questions about langfuse-observability

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

FAQPage Schema
How do I implement LLM observability for multi-agent workflows?▼

You implement LLM observability by initializing the Langfuse client and applying the @observe decorator to LLM calls, which automatically captures parent-child distributed tracing for multi-agent workflows.

What is distributed tracing for LLM applications and how does it work?▼

Distributed tracing for LLM applications works by generating automatic parent-child spans across multi-agent workflows, linking token usage and versioned prompts to traces for full production visibility.

Can I track token cost calculation per model in production LLM applications?▼

Yes, you can track cost calculation per model by capturing token usage data from LLM calls and applying per-model pricing to visualize expenses on integrated cost tracking dashboards.

Does prompt management support versioning linked to LLM traces?▼

Yes, prompt management supports versioned prompts linked directly to traces, ensuring reproducibility by tying specific prompt versions to their corresponding LLM execution spans.

What is the best way to evaluate LLM quality trends in production?▼

The best way to evaluate LLM quality trends is by assigning evaluation scores to traces and visualizing quality changes over time on integrated dashboards grouped by session or user.

Do I need to group traces by session to track user journeys?▼

Yes, you need to group traces by session or user to track journeys, allowing the observability platform to aggregate token usage, cost tracking, and evaluation scores per user session.

Related Skills

Explore More Software Engineering Skills→
ddnettersddnetters
2

langfuse-observability

Query Langfuse REST API for LLM traces, generations, costs, and metrics.

Community
Intermediate
ice-droid-99ice-droid-99

querying-mlflow-metrics

Fetch aggregated MLflow trace metrics for token usage, latency, and quality scores.

Community
Intermediate
faisalanjumfaisalanjum
3

test-workflow-parent

Test parent skill continuation after child skill execution completes.

Community
Intermediate
Sheldon-92Sheldon-92
3

llm-observability

Trace LLM calls with OpenTelemetry GenAI semantic conventions and enforce governance policies.

Community
Advanced