Langfuse

Use this skill to integrate codebases and workflows with Langfuse, the open-source LLM engineering platform for tracing, prompt management, and evaluation.

Treat self-hosted Langfuse as a first-class deployment mode, not a special case. When the user mentions private infrastructure, on-prem, internal URLs, custom domains, or self-hosting, configure examples and guidance around their deployment URL and platform compatibility.

Prefer SDK-based examples for Python and JS/TS. Use the public API when the task is language-agnostic, needs direct HTTP examples, or fits an existing automation pipeline.

Quick Decision Tree

1. Migrating hardcoded prompts into Langfuse?

   • Read references/prompt-migration.md.
   • Create prompts in Langfuse, replace inline prompt literals with fetch/render calls, and preserve variable semantics.

2. Adding observability/tracing to an app or agent?

   • Read references/tracing-and-querying.md.
   • Prefer native SDK instrumentation and OpenTelemetry-compatible patterns over bespoke logging.