What problem does it solve? Working with Langfuse requires up-to-date knowledge of its frequently changing APIs, SDKs, and documentation, and manually querying traces, prompts, datasets, or scores through the UI is slow and hard to automate. ## Core Features & Use Cases - CLI-Based API Access: Query and modify any Langfuse resource (traces, prompts, datasets, scores, sessions) programmatically using langfuse-cli via npx, with schema discovery through api __schema. - Documentation Retrieval: Access current Langfuse docs through llms.txt, markdown page fetches, and a search API covering docs plus GitHub issues and discussions. - Use-Case Reference Guides: Follow dedicated guides for instrumentation, prompt migration, prompt engineering, user feedback capture, error analysis, judge calibration, CI/CD experiment gating, evaluator upgrades, and v4 migration. - Use Case: When debugging why an LLM application produces bad outputs, use the error-analysis reference to sample traces, build a failure taxonomy, create annotation queues, and decide what to fix. ## Quick Start Set the LANGFUSE_PUBLIC_KEY, LANGFUSE_SECRET_KEY, and LANGFUSE_BASE_URL environment variables, then ask the assistant to list recent traces or look up a Langfuse documentation topic.
