langfuse - Use Langfuse for LLM observability and evaluation
Supports LLM tracing, prompt management, evaluation, dataset handling, cost and performance monitoring, and provenance-aware workflow execution with Langfuse.
Tags
Updated: 2026-10-06Capabilities
Typical Inputs
Typical Outputs
What this skill does
- Trace LLM applications
- Manage prompt versions
- Evaluate LLM outputs
- Manage evaluation datasets
- Track application costs
- Monitor performance metrics
- Test prompt variants
- Preserve source provenance
Inputs
- User task scope
- Copied upstream files
- Repository context
- Provenance metadata
- Langfuse account
- LLM API keys
Outputs
- LLM traces
- Prompt versions
- Evaluation results
- Evaluation datasets
- Cost metrics
- Performance metrics
- Provenance records
- Review summaries
Requirements
- Langfuse account
- LLM API keys
- Python or TypeScript/JavaScript
- Infrastructure for self-hosting