Index

NexusLabs.Needlr.AgentFramework.Langfuse Namespace

Classes
LangfuseDatasetItem	A single Langfuse dataset item — one eval case. Upserted via UpsertItemAsync(LangfuseDatasetItem, CancellationToken) and later referenced by an experiment run.
LangfuseEvaluationScoreExtensions	Convenience extensions for projecting Microsoft.Extensions.AI.Evaluation results onto a Langfuse scenario trace as scores.
LangfuseException	The exception thrown when a request to the Langfuse API fails.
LangfuseMetric	One metric in a LangfuseMetricsQuery: a measure plus an aggregation. The result column is named {Aggregation}_{Measure} (for example avg_value, sum_totalCost).
LangfuseMetricsFilter	One filter in a LangfuseMetricsQuery.
LangfuseMetricsQuery	A query against the Langfuse Metrics API (GET /api/public/metrics). Use it to read aggregates back from Langfuse — for example to drive a CI quality gate from recorded eval scores.
LangfuseMetricsResult	The result of a LangfuseMetricsQuery. Each row is a dictionary of column name to value; aggregated columns are named {Aggregation}_{Measure} and dimension columns keep their field name.
LangfuseModelPrice	A model price definition registered with Langfuse so it can compute cost for generations whose model it does not price out of the box (for example mock or internal model names). Langfuse derives cost from token usage and these prices when a generation carries no explicit cost.
LangfuseOptions	Configuration for exporting Needlr agent telemetry and evaluation scores to Langfuse.
LangfusePrompt	A prompt fetched from Langfuse prompt management. Use Text for a text prompt's content, and pass the prompt to SetPrompt(LangfusePrompt) to link generations to the exact version that was used.
LangfuseScoreConfig	Declares the shape of a named Langfuse score so the dashboard renders consistent ranges, colors, and category sets and validates incoming values. Created via EnsureScoreConfigAsync(LangfuseScoreConfig, CancellationToken).
LangfuseScoreConfigCategory	One allowed category for a Categorical score config.
LangfuseScoreError	Describes a single failed attempt to upload an evaluation score to Langfuse. Passed to ScoreErrorCallback when NonFatal is in effect, so callers can log the loss with their own logger.
LangfuseServiceCollectionExtensions	Registers Langfuse OTLP export on an Microsoft.Extensions.DependencyInjection.IServiceCollection for ASP.NET Core and generic-host applications.
LangfuseTelemetry	Entry point for exporting Needlr agent telemetry to Langfuse without requiring a generic host. Designed for evals, console apps, and test fixtures that build telemetry by hand.

Interfaces
ILangfuseDatasetClient	Creates and populates Langfuse datasets — the named collections of eval cases that experiment runs are scored against.
ILangfuseExperimentRun	Represents an in-progress Langfuse experiment (dataset run). Each call to BeginItemAsync(string, string, IEnumerable<string>, IReadOnlyDictionary<string,string>, CancellationToken) starts a trace for one dataset item and links it to this run, so the scores recorded on that trace roll up into the run's experiment-comparison view.
ILangfuseMetricsClient	Reads aggregates back from Langfuse via the Metrics API, so recorded eval scores (and observation metrics) can drive CI quality gates and dashboards from Langfuse as the source of truth.
ILangfuseModelClient	Registers model price definitions with Langfuse so it can compute cost for generations whose model it does not price by default.
ILangfusePromptClient	Fetches and creates prompts in Langfuse prompt management, so the prompt an eval runs against lives in Langfuse and the generations link to the exact version that was used.
ILangfuseScenario	Represents a single Langfuse trace scoped to one eval scenario or agent run. Created via BeginScenario(string, string, string, IEnumerable<string>, IReadOnlyDictionary<string,string>).
ILangfuseScoreClient	Records Langfuse scores against a known trace id. Registered in dependency injection by AddNeedlrLangfuse(this IServiceCollection, Action<LangfuseOptions>) for ASP.NET Core and generic-host applications that score their own request traces.
ILangfuseScoreConfigClient	Idempotently registers Langfuse score configs so eval scores appear on the dashboard with defined data types, ranges, and category sets.
ILangfuseSession	Represents an active Langfuse export session. Owns the OpenTelemetry tracer/meter providers that forward Needlr agent telemetry to Langfuse and flushes pending telemetry on disposal.

Enums
LangfuseMetricsView	The Langfuse Metrics API view to query. Determines which measures and dimensions are available.
LangfuseRegion	Identifies the Langfuse deployment that traces and scores are sent to. Used to derive the base URL of the OTLP ingestion endpoint when Host is not set explicitly.
LangfuseScoreDataType	The data type of a Langfuse score, controlling how the score value is interpreted and displayed.
LangfuseScoreFailureMode	Controls how a Langfuse scenario reacts when uploading an evaluation score to the Langfuse Scores API fails.