langfuse

Trace, evaluate, and manage prompts for LLM applications on Langfuse.

10|2|Updated Mar 10, 2026

Search Tags:#debugging #prompt engineering #llm #evaluation #observability #tracing

Ready to run

npx skills add https://github.com/bugrabilge/bilge-development-kit --skill langfuse-bugrabilge

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: langfuse
Source: https://github.com/bugrabilge/bilge-development-kit/tree/main/skills-extra/langfuse
Command: npx skills add https://github.com/bugrabilge/bilge-development-kit --skill langfuse-bugrabilge

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

This Skill provides deep visibility into your LLM applications, enabling effective debugging, performance monitoring, and prompt optimization through comprehensive tracing and evaluation.

Core Features & Use Cases

LLM Tracing: Monitor every LLM call, including inputs, outputs, and latency.
Prompt Management: Version, test, and manage your prompts efficiently.
Evaluation: Score LLM outputs against defined metrics and datasets.
Use Case: Debug a customer support chatbot by tracing conversations to identify where responses become unhelpful or slow, then use prompt management to iterate on better prompts.

Quick Start

Use the langfuse skill to instrument your OpenAI API calls for tracing.

Frequently Asked Questions about langfuse

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

FAQPage Schema

How do I trace LLM API calls in my Python application for debugging?▼

LLM tracing monitors every LLM call, capturing inputs, outputs, and latency. You can instrument your OpenAI API calls to trace conversations and identify where responses become unhelpful or slow.

What is LLM observability and when do I need it?▼

LLM observability provides deep visibility into your LLM applications. You need it when debugging complex interactions, monitoring performance, or optimizing prompts to ensure effective and reliable model responses.

Can I use this for prompt management and evaluation in TypeScript?▼

Prompt management and evaluation are supported for both Python and TypeScript applications. You can version, test, manage prompts, and score LLM outputs against defined metrics and datasets.

Does LLM tracing work with LangChain and LlamaIndex integrations?▼

Tracing integrates with LLM APIs and frameworks like OpenAI, LangChain, and LlamaIndex. This allows you to monitor calls across different platforms within your existing application stack.

Do I need a Langfuse account to start tracing my LLM calls?▼

A Langfuse account and API keys are required for integration. This setup enables the platform to facilitate tracing, prompt management, evaluation, and dataset management for your applications.

What's the best way to debug a customer support chatbot's unhelpful responses?▼

Trace conversations to identify where responses become unhelpful or slow, then use prompt management to iterate on better prompts. This targets the exact failure points in the chatbot flow.