Langfuse MCP Server

import { z } from 'zod'; import { LangfuseAnalyticsClient } from '../langfuse-client.js'; import { ModelUsage } from '../types.js'; export const usageByModelSchema = z.object({ from: z.string().datetime(), to: z.string().datetime(), environment: z.string().optional(), limit: z.number().optional().default(20), }); export async function usageByModel( client: LangfuseAnalyticsClient, args: z.infer<typeof usageByModelSchema> ) { // Use daily metrics API which works correctly (same approach as get-cost-analysis) try { const dailyResponse = await client.getDailyMetrics({ tags: args.environment ? [`environment:${args.environment}`] : undefined, }); const modelMap = new Map<string, { totalCost: number; totalTokens: number; observationCount: number; }>(); if (dailyResponse.data && Array.isArray(dailyResponse.data)) { // Filter by date range and aggregate model data const fromDate = new Date(args.from); const toDate = new Date(args.to); dailyResponse.data.forEach((day: any) => { const dayDate = new Date(day.date); if (dayDate >= fromDate && dayDate <= toDate) { // Process model usage from daily breakdown if (day.usage && Array.isArray(day.usage)) { day.usage.forEach((usage: any) => { const modelName = usage.model || 'unknown'; const existing = modelMap.get(modelName) || { totalCost: 0, totalTokens: 0, observationCount: 0 }; modelMap.set(modelName, { totalCost: existing.totalCost + (usage.totalCost || 0), totalTokens: existing.totalTokens + (usage.totalUsage || usage.inputUsage + usage.outputUsage || 0), observationCount: existing.observationCount + (usage.countObservations || 0), }); }); } } }); } // Convert map to array and sort by cost const modelUsages: ModelUsage[] = Array.from(modelMap.entries()).map(([model, data]) => ({ model, totalCost: data.totalCost, totalTokens: data.totalTokens, observationCount: data.observationCount, })); // Sort by cost descending and limit results const sortedModels = modelUsages .sort((a, b) => b.totalCost - a.totalCost) .slice(0, args.limit); return { content: [ { type: 'text' as const, text: JSON.stringify( { projectId: client.getProjectId(), from: args.from, to: args.to, models: sortedModels, }, null, 2 ), }, ], }; } catch (error) { // Fallback to broken metrics API for debugging console.error('Error using daily metrics for usage by model:', error); const filters: any[] = []; if (args.environment) { filters.push({ column: 'environment', operator: 'equals', value: args.environment, type: 'string', }); } const response = await client.getMetrics({ view: 'observations', from: args.from, to: args.to, metrics: [ { measure: 'totalCost', aggregation: 'sum' }, { measure: 'totalTokens', aggregation: 'sum' }, { measure: 'count', aggregation: 'count' }, ], dimensions: [{ field: 'providedModelName' }], filters, }); console.log('DEBUG - Fallback metrics response:', JSON.stringify(response, null, 2)); const modelUsages: ModelUsage[] = []; // Parse the response data - use correct aggregated field names if (response.data && Array.isArray(response.data)) { response.data.forEach((row: any) => { modelUsages.push({ model: row.providedModelName || 'unknown', totalCost: row.totalCost_sum || 0, totalTokens: row.totalTokens_sum || 0, observationCount: row.count_count || 0, }); }); } // Sort by cost descending and limit results const sortedModels = modelUsages .sort((a, b) => b.totalCost - a.totalCost) .slice(0, args.limit); return { content: [ { type: 'text' as const, text: JSON.stringify( { projectId: client.getProjectId(), from: args.from, to: args.to, models: sortedModels, error: 'Used fallback metrics API due to daily metrics error', }, null, 2 ), }, ], }; } }