toad-eye 1.1.0 → 2.1.0

package/templates/grafana/dashboards/latency-analysis.json

@@ -0,0 +1,134 @@
+{
+  "annotations": { "list": [] },
+  "editable": true,
+  "fiscalYearStartMonth": 0,
+  "graphTooltip": 1,
+  "id": null,
+  "links": [],
+  "panels": [
+    {
+      "title": "Latency p50 / p95 / p99 by Model",
+      "type": "timeseries",
+      "gridPos": { "h": 8, "w": 24, "x": 0, "y": 0 },
+      "datasource": { "type": "prometheus", "uid": "prometheus" },
+      "targets": [
+        {
+          "expr": "histogram_quantile(0.50, sum(rate(gen_ai_client_operation_duration_milliseconds_bucket{gen_ai_provider_name=~\"$provider\", gen_ai_request_model=~\"$model\"}[5m])) by (le, gen_ai_request_model))",
+          "legendFormat": "p50 {{ gen_ai_request_model }}",
+          "refId": "A"
+        },
+        {
+          "expr": "histogram_quantile(0.95, sum(rate(gen_ai_client_operation_duration_milliseconds_bucket{gen_ai_provider_name=~\"$provider\", gen_ai_request_model=~\"$model\"}[5m])) by (le, gen_ai_request_model))",
+          "legendFormat": "p95 {{ gen_ai_request_model }}",
+          "refId": "B"
+        },
+        {
+          "expr": "histogram_quantile(0.99, sum(rate(gen_ai_client_operation_duration_milliseconds_bucket{gen_ai_provider_name=~\"$provider\", gen_ai_request_model=~\"$model\"}[5m])) by (le, gen_ai_request_model))",
+          "legendFormat": "p99 {{ gen_ai_request_model }}",
+          "refId": "C"
+        }
+      ],
+      "fieldConfig": {
+        "defaults": {
+          "unit": "ms",
+          "custom": { "drawStyle": "line", "fillOpacity": 15, "lineWidth": 2 }
+        },
+        "overrides": []
+      }
+    },
+    {
+      "title": "Latency Distribution",
+      "type": "histogram",
+      "gridPos": { "h": 8, "w": 12, "x": 0, "y": 8 },
+      "datasource": { "type": "prometheus", "uid": "prometheus" },
+      "targets": [
+        {
+          "expr": "sum(rate(gen_ai_client_operation_duration_milliseconds_bucket{gen_ai_provider_name=~\"$provider\", gen_ai_request_model=~\"$model\"}[5m])) by (le)",
+          "legendFormat": "{{ le }}",
+          "refId": "A",
+          "format": "heatmap"
+        }
+      ],
+      "fieldConfig": {
+        "defaults": {
+          "unit": "ms",
+          "custom": { "fillOpacity": 50 }
+        },
+        "overrides": []
+      }
+    },
+    {
+      "title": "Avg Latency by Model",
+      "type": "timeseries",
+      "gridPos": { "h": 8, "w": 12, "x": 12, "y": 8 },
+      "datasource": { "type": "prometheus", "uid": "prometheus" },
+      "targets": [
+        {
+          "expr": "sum(rate(gen_ai_client_operation_duration_milliseconds_sum{gen_ai_provider_name=~\"$provider\", gen_ai_request_model=~\"$model\"}[5m])) by (gen_ai_request_model) / sum(rate(gen_ai_client_operation_duration_milliseconds_count{gen_ai_provider_name=~\"$provider\", gen_ai_request_model=~\"$model\"}[5m])) by (gen_ai_request_model)",
+          "legendFormat": "{{ gen_ai_request_model }}",
+          "refId": "A"
+        }
+      ],
+      "fieldConfig": {
+        "defaults": {
+          "unit": "ms",
+          "custom": { "drawStyle": "line", "fillOpacity": 15, "lineWidth": 2 }
+        },
+        "overrides": []
+      }
+    },
+    {
+      "title": "Slow Requests p99 by Provider",
+      "type": "timeseries",
+      "gridPos": { "h": 8, "w": 24, "x": 0, "y": 16 },
+      "datasource": { "type": "prometheus", "uid": "prometheus" },
+      "targets": [
+        {
+          "expr": "histogram_quantile(0.99, sum(rate(gen_ai_client_operation_duration_milliseconds_bucket{gen_ai_provider_name=~\"$provider\", gen_ai_request_model=~\"$model\"}[5m])) by (le, gen_ai_provider_name))",
+          "legendFormat": "p99 {{ gen_ai_provider_name }}",
+          "refId": "A"
+        }
+      ],
+      "fieldConfig": {
+        "defaults": {
+          "unit": "ms",
+          "custom": { "drawStyle": "line", "fillOpacity": 15, "lineWidth": 2 }
+        },
+        "overrides": []
+      }
+    }
+  ],
+  "refresh": "5s",
+  "schemaVersion": 39,
+  "tags": ["llm", "observability", "toad-eye"],
+  "templating": {
+    "list": [
+      {
+        "name": "provider",
+        "type": "query",
+        "datasource": { "type": "prometheus", "uid": "prometheus" },
+        "query": "label_values(gen_ai_client_requests_total, gen_ai_provider_name)",
+        "includeAll": true,
+        "allValue": ".*",
+        "current": { "text": "All", "value": "$__all" },
+        "multi": true
+      },
+      {
+        "name": "model",
+        "type": "query",
+        "datasource": { "type": "prometheus", "uid": "prometheus" },
+        "query": "label_values(gen_ai_client_requests_total{gen_ai_provider_name=~\"$provider\"}, gen_ai_request_model)",
+        "includeAll": true,
+        "allValue": ".*",
+        "current": { "text": "All", "value": "$__all" },
+        "multi": true
+      }
+    ]
+  },
+  "time": { "from": "now-15m", "to": "now" },
+  "timepicker": {},
+  "timezone": "browser",
+  "title": "toad-eye: Latency Analysis",
+  "uid": "toad-eye-latency",
+  "version": 1
+}

package/templates/grafana/dashboards/llm-overview.json

@@ -7,14 +7,115 @@
   "links": [],
   "panels": [
     {
-      "title": "Request rate (per minute)",
+      "title": "Total Requests",
+      "type": "stat",
+      "gridPos": { "h": 4, "w": 6, "x": 0, "y": 0 },
+      "datasource": { "type": "prometheus", "uid": "prometheus" },
+      "targets": [
+        {
+          "expr": "sum(gen_ai_client_requests_total{gen_ai_provider_name=~\"$provider\", gen_ai_request_model=~\"$model\"})",
+          "refId": "A"
+        }
+      ],
+      "fieldConfig": {
+        "defaults": {
+          "unit": "short",
+          "thresholds": {
+            "steps": [
+              { "color": "green", "value": null },
+              { "color": "yellow", "value": 1000 },
+              { "color": "red", "value": 10000 }
+            ]
+          }
+        },
+        "overrides": []
+      }
+    },
+    {
+      "title": "Error Rate %",
+      "type": "stat",
+      "gridPos": { "h": 4, "w": 6, "x": 6, "y": 0 },
+      "datasource": { "type": "prometheus", "uid": "prometheus" },
+      "targets": [
+        {
+          "expr": "sum(gen_ai_client_errors_total{gen_ai_provider_name=~\"$provider\", gen_ai_request_model=~\"$model\"}) / sum(gen_ai_client_requests_total{gen_ai_provider_name=~\"$provider\", gen_ai_request_model=~\"$model\"}) * 100",
+          "refId": "A"
+        }
+      ],
+      "fieldConfig": {
+        "defaults": {
+          "unit": "percent",
+          "thresholds": {
+            "steps": [
+              { "color": "green", "value": null },
+              { "color": "yellow", "value": 5 },
+              { "color": "red", "value": 20 }
+            ]
+          }
+        },
+        "overrides": []
+      }
+    },
+    {
+      "title": "Avg Latency",
+      "type": "stat",
+      "gridPos": { "h": 4, "w": 6, "x": 12, "y": 0 },
+      "datasource": { "type": "prometheus", "uid": "prometheus" },
+      "targets": [
+        {
+          "expr": "sum(rate(gen_ai_client_operation_duration_milliseconds_sum{gen_ai_provider_name=~\"$provider\", gen_ai_request_model=~\"$model\"}[5m])) / sum(rate(gen_ai_client_operation_duration_milliseconds_count{gen_ai_provider_name=~\"$provider\", gen_ai_request_model=~\"$model\"}[5m]))",
+          "refId": "A"
+        }
+      ],
+      "fieldConfig": {
+        "defaults": {
+          "unit": "ms",
+          "thresholds": {
+            "steps": [
+              { "color": "green", "value": null },
+              { "color": "yellow", "value": 1000 },
+              { "color": "red", "value": 5000 }
+            ]
+          }
+        },
+        "overrides": []
+      }
+    },
+    {
+      "title": "Total Cost",
+      "type": "stat",
+      "gridPos": { "h": 4, "w": 6, "x": 18, "y": 0 },
+      "datasource": { "type": "prometheus", "uid": "prometheus" },
+      "targets": [
+        {
+          "expr": "sum(gen_ai_client_request_cost_USD_sum{gen_ai_provider_name=~\"$provider\", gen_ai_request_model=~\"$model\"})",
+          "refId": "A"
+        }
+      ],
+      "fieldConfig": {
+        "defaults": {
+          "unit": "currencyUSD",
+          "thresholds": {
+            "steps": [
+              { "color": "green", "value": null },
+              { "color": "yellow", "value": 10 },
+              { "color": "red", "value": 100 }
+            ]
+          }
+        },
+        "overrides": []
+      }
+    },
+    {
+      "title": "Request Rate per minute",
       "type": "timeseries",
-      "gridPos": { "h": 8, "w": 12, "x": 0, "y": 0 },
+      "gridPos": { "h": 8, "w": 12, "x": 0, "y": 4 },
       "datasource": { "type": "prometheus", "uid": "prometheus" },
       "targets": [
         {
-          "expr": "sum(rate(llm_requests_total[1m])) by (provider) * 60",
-          "legendFormat": "{{ provider }}"
+          "expr": "sum(rate(gen_ai_client_requests_total{gen_ai_provider_name=~\"$provider\", gen_ai_request_model=~\"$model\"}[1m])) by (gen_ai_provider_name) * 60",
+          "legendFormat": "{{ gen_ai_provider_name }}",
+          "refId": "A"
         }
       ],
       "fieldConfig": {
@@ -26,14 +127,15 @@
       }
     },
     {
-      "title": "Error rate (per minute)",
+      "title": "Error Rate per minute",
       "type": "timeseries",
-      "gridPos": { "h": 8, "w": 12, "x": 12, "y": 0 },
+      "gridPos": { "h": 8, "w": 12, "x": 12, "y": 4 },
       "datasource": { "type": "prometheus", "uid": "prometheus" },
       "targets": [
         {
-          "expr": "sum(rate(llm_errors_total[1m])) by (provider) * 60",
-          "legendFormat": "{{ provider }}"
+          "expr": "sum(rate(gen_ai_client_errors_total{gen_ai_provider_name=~\"$provider\", gen_ai_request_model=~\"$model\"}[1m])) by (gen_ai_provider_name) * 60",
+          "legendFormat": "{{ gen_ai_provider_name }}",
+          "refId": "A"
         }
       ],
       "fieldConfig": {
@@ -48,22 +150,24 @@
     {
       "title": "Latency p50 / p95 (ms)",
       "type": "timeseries",
-      "gridPos": { "h": 8, "w": 12, "x": 0, "y": 8 },
+      "gridPos": { "h": 8, "w": 12, "x": 0, "y": 12 },
       "datasource": { "type": "prometheus", "uid": "prometheus" },
       "targets": [
         {
-          "expr": "histogram_quantile(0.50, sum(rate(llm_request_duration_milliseconds_bucket[1m])) by (le, provider))",
-          "legendFormat": "p50 {{ provider }}"
+          "expr": "histogram_quantile(0.50, sum(rate(gen_ai_client_operation_duration_milliseconds_bucket{gen_ai_provider_name=~\"$provider\", gen_ai_request_model=~\"$model\"}[5m])) by (le))",
+          "legendFormat": "p50",
+          "refId": "A"
         },
         {
-          "expr": "histogram_quantile(0.95, sum(rate(llm_request_duration_milliseconds_bucket[1m])) by (le, provider))",
-          "legendFormat": "p95 {{ provider }}"
+          "expr": "histogram_quantile(0.95, sum(rate(gen_ai_client_operation_duration_milliseconds_bucket{gen_ai_provider_name=~\"$provider\", gen_ai_request_model=~\"$model\"}[5m])) by (le))",
+          "legendFormat": "p95",
+          "refId": "B"
         }
       ],
       "fieldConfig": {
         "defaults": {
           "unit": "ms",
-          "custom": { "drawStyle": "line", "fillOpacity": 10, "lineWidth": 2 }
+          "custom": { "drawStyle": "line", "fillOpacity": 15, "lineWidth": 2 }
         },
         "overrides": []
       }
@@ -71,12 +175,13 @@
     {
       "title": "Cost per minute (USD)",
       "type": "timeseries",
-      "gridPos": { "h": 8, "w": 12, "x": 12, "y": 8 },
+      "gridPos": { "h": 8, "w": 12, "x": 12, "y": 12 },
       "datasource": { "type": "prometheus", "uid": "prometheus" },
       "targets": [
         {
-          "expr": "sum(rate(llm_request_cost_USD_sum[1m])) by (provider) * 60",
-          "legendFormat": "{{ provider }}"
+          "expr": "sum(rate(gen_ai_client_request_cost_USD_sum{gen_ai_provider_name=~\"$provider\", gen_ai_request_model=~\"$model\"}[1m])) by (gen_ai_provider_name) * 60",
+          "legendFormat": "{{ gen_ai_provider_name }}",
+          "refId": "A"
         }
       ],
       "fieldConfig": {
@@ -88,13 +193,14 @@
       }
     },
     {
-      "title": "Total tokens consumed",
+      "title": "Total Tokens",
       "type": "stat",
-      "gridPos": { "h": 4, "w": 8, "x": 0, "y": 16 },
+      "gridPos": { "h": 4, "w": 8, "x": 0, "y": 20 },
       "datasource": { "type": "prometheus", "uid": "prometheus" },
       "targets": [
         {
-          "expr": "sum(llm_tokens_total)"
+          "expr": "sum(gen_ai_client_token_usage_total{gen_ai_provider_name=~\"$provider\", gen_ai_request_model=~\"$model\"})",
+          "refId": "A"
         }
       ],
       "fieldConfig": {
@@ -112,13 +218,14 @@
       }
     },
     {
-      "title": "Total requests",
+      "title": "Total Requests",
       "type": "stat",
-      "gridPos": { "h": 4, "w": 8, "x": 8, "y": 16 },
+      "gridPos": { "h": 4, "w": 8, "x": 8, "y": 20 },
       "datasource": { "type": "prometheus", "uid": "prometheus" },
       "targets": [
         {
-          "expr": "sum(llm_requests_total)"
+          "expr": "sum(gen_ai_client_requests_total{gen_ai_provider_name=~\"$provider\", gen_ai_request_model=~\"$model\"})",
+          "refId": "A"
         }
       ],
       "fieldConfig": {
@@ -136,13 +243,14 @@
       }
     },
     {
-      "title": "Total errors",
+      "title": "Total Errors",
       "type": "stat",
-      "gridPos": { "h": 4, "w": 8, "x": 16, "y": 16 },
+      "gridPos": { "h": 4, "w": 8, "x": 16, "y": 20 },
       "datasource": { "type": "prometheus", "uid": "prometheus" },
       "targets": [
         {
-          "expr": "sum(llm_errors_total)"
+          "expr": "sum(gen_ai_client_errors_total{gen_ai_provider_name=~\"$provider\", gen_ai_request_model=~\"$model\"})",
+          "refId": "A"
         }
       ],
       "fieldConfig": {
@@ -163,11 +271,34 @@
   "refresh": "5s",
   "schemaVersion": 39,
   "tags": ["llm", "observability", "toad-eye"],
-  "templating": { "list": [] },
+  "templating": {
+    "list": [
+      {
+        "name": "provider",
+        "type": "query",
+        "datasource": { "type": "prometheus", "uid": "prometheus" },
+        "query": "label_values(gen_ai_client_requests_total, gen_ai_provider_name)",
+        "includeAll": true,
+        "allValue": ".*",
+        "current": { "text": "All", "value": "$__all" },
+        "multi": true
+      },
+      {
+        "name": "model",
+        "type": "query",
+        "datasource": { "type": "prometheus", "uid": "prometheus" },
+        "query": "label_values(gen_ai_client_requests_total{gen_ai_provider_name=~\"$provider\"}, gen_ai_request_model)",
+        "includeAll": true,
+        "allValue": ".*",
+        "current": { "text": "All", "value": "$__all" },
+        "multi": true
+      }
+    ]
+  },
   "time": { "from": "now-15m", "to": "now" },
   "timepicker": {},
   "timezone": "browser",
-  "title": "toad-eye: LLM observability",
-  "uid": "toad-eye-llm-overview",
-  "version": 2
+  "title": "toad-eye: Overview",
+  "uid": "toad-eye-overview",
+  "version": 1
 }

package/templates/grafana/dashboards/model-comparison.json

@@ -0,0 +1,143 @@
+{
+  "annotations": { "list": [] },
+  "editable": true,
+  "fiscalYearStartMonth": 0,
+  "graphTooltip": 1,
+  "id": null,
+  "links": [],
+  "panels": [
+    {
+      "title": "Latency by Model (p95)",
+      "type": "timeseries",
+      "gridPos": { "h": 8, "w": 12, "x": 0, "y": 0 },
+      "datasource": { "type": "prometheus", "uid": "prometheus" },
+      "targets": [
+        {
+          "expr": "histogram_quantile(0.95, sum(rate(gen_ai_client_operation_duration_milliseconds_bucket{gen_ai_provider_name=~\"$provider\", gen_ai_request_model=~\"$model\"}[5m])) by (le, gen_ai_request_model))",
+          "legendFormat": "p95 {{ gen_ai_request_model }}",
+          "refId": "A"
+        }
+      ],
+      "fieldConfig": {
+        "defaults": {
+          "unit": "ms",
+          "custom": { "drawStyle": "line", "fillOpacity": 15, "lineWidth": 2 }
+        },
+        "overrides": []
+      }
+    },
+    {
+      "title": "Cost by Model (avg per request)",
+      "type": "timeseries",
+      "gridPos": { "h": 8, "w": 12, "x": 12, "y": 0 },
+      "datasource": { "type": "prometheus", "uid": "prometheus" },
+      "targets": [
+        {
+          "expr": "sum(rate(gen_ai_client_request_cost_USD_sum{gen_ai_provider_name=~\"$provider\", gen_ai_request_model=~\"$model\"}[5m])) by (gen_ai_request_model) / sum(rate(gen_ai_client_request_cost_USD_count{gen_ai_provider_name=~\"$provider\", gen_ai_request_model=~\"$model\"}[5m])) by (gen_ai_request_model)",
+          "legendFormat": "{{ gen_ai_request_model }}",
+          "refId": "A"
+        }
+      ],
+      "fieldConfig": {
+        "defaults": {
+          "unit": "currencyUSD",
+          "custom": { "drawStyle": "line", "fillOpacity": 15, "lineWidth": 2 }
+        },
+        "overrides": []
+      }
+    },
+    {
+      "title": "Error Rate by Model (%)",
+      "type": "timeseries",
+      "gridPos": { "h": 8, "w": 12, "x": 0, "y": 8 },
+      "datasource": { "type": "prometheus", "uid": "prometheus" },
+      "targets": [
+        {
+          "expr": "sum(rate(gen_ai_client_errors_total{gen_ai_provider_name=~\"$provider\", gen_ai_request_model=~\"$model\"}[5m])) by (gen_ai_request_model) / sum(rate(gen_ai_client_requests_total{gen_ai_provider_name=~\"$provider\", gen_ai_request_model=~\"$model\"}[5m])) by (gen_ai_request_model) * 100",
+          "legendFormat": "{{ gen_ai_request_model }}",
+          "refId": "A"
+        }
+      ],
+      "fieldConfig": {
+        "defaults": {
+          "unit": "percent",
+          "custom": { "drawStyle": "line", "fillOpacity": 15, "lineWidth": 2 }
+        },
+        "overrides": []
+      }
+    },
+    {
+      "title": "Throughput by Model (req/min)",
+      "type": "timeseries",
+      "gridPos": { "h": 8, "w": 12, "x": 12, "y": 8 },
+      "datasource": { "type": "prometheus", "uid": "prometheus" },
+      "targets": [
+        {
+          "expr": "sum(rate(gen_ai_client_requests_total{gen_ai_provider_name=~\"$provider\", gen_ai_request_model=~\"$model\"}[1m])) by (gen_ai_request_model) * 60",
+          "legendFormat": "{{ gen_ai_request_model }}",
+          "refId": "A"
+        }
+      ],
+      "fieldConfig": {
+        "defaults": {
+          "unit": "reqpm",
+          "custom": { "drawStyle": "line", "fillOpacity": 15, "lineWidth": 2 }
+        },
+        "overrides": []
+      }
+    },
+    {
+      "title": "Tokens per Request by Model",
+      "type": "timeseries",
+      "gridPos": { "h": 8, "w": 24, "x": 0, "y": 16 },
+      "datasource": { "type": "prometheus", "uid": "prometheus" },
+      "targets": [
+        {
+          "expr": "sum(rate(gen_ai_client_token_usage_total{gen_ai_provider_name=~\"$provider\", gen_ai_request_model=~\"$model\"}[5m])) by (gen_ai_request_model) / sum(rate(gen_ai_client_requests_total{gen_ai_provider_name=~\"$provider\", gen_ai_request_model=~\"$model\"}[5m])) by (gen_ai_request_model)",
+          "legendFormat": "{{ gen_ai_request_model }}",
+          "refId": "A"
+        }
+      ],
+      "fieldConfig": {
+        "defaults": {
+          "unit": "short",
+          "custom": { "drawStyle": "line", "fillOpacity": 15, "lineWidth": 2 }
+        },
+        "overrides": []
+      }
+    }
+  ],
+  "refresh": "5s",
+  "schemaVersion": 39,
+  "tags": ["llm", "observability", "toad-eye"],
+  "templating": {
+    "list": [
+      {
+        "name": "provider",
+        "type": "query",
+        "datasource": { "type": "prometheus", "uid": "prometheus" },
+        "query": "label_values(gen_ai_client_requests_total, gen_ai_provider_name)",
+        "includeAll": true,
+        "allValue": ".*",
+        "current": { "text": "All", "value": "$__all" },
+        "multi": true
+      },
+      {
+        "name": "model",
+        "type": "query",
+        "datasource": { "type": "prometheus", "uid": "prometheus" },
+        "query": "label_values(gen_ai_client_requests_total{gen_ai_provider_name=~\"$provider\"}, gen_ai_request_model)",
+        "includeAll": true,
+        "allValue": ".*",
+        "current": { "text": "All", "value": "$__all" },
+        "multi": true
+      }
+    ]
+  },
+  "time": { "from": "now-15m", "to": "now" },
+  "timepicker": {},
+  "timezone": "browser",
+  "title": "toad-eye: Model Comparison",
+  "uid": "toad-eye-comparison",
+  "version": 1
+}

package/dist/metrics.d.ts

@@ -1,6 +0,0 @@
-export declare function initMetrics(): void;
-export declare function recordRequestDuration(ms: number, provider: string, model: string): void;
-export declare function recordRequestCost(usd: number, provider: string, model: string): void;
-export declare function recordTokens(count: number, provider: string, model: string): void;
-export declare function recordRequest(provider: string, model: string): void;
-export declare function recordError(provider: string, model: string): void;

package/dist/metrics.js

@@ -1,46 +0,0 @@
-import { metrics } from "@opentelemetry/api";
-import { LLM_METRICS, INSTRUMENTATION_NAME } from "./types.js";
-let requestDuration;
-let requestCost;
-let tokensTotal;
-let requestsTotal;
-let errorsTotal;
-let initialized = false;
-export function initMetrics() {
-    if (initialized)
-        return;
-    const meter = metrics.getMeter(INSTRUMENTATION_NAME);
-    requestDuration = meter.createHistogram(LLM_METRICS.REQUEST_DURATION, {
-        description: "LLM request duration in milliseconds",
-        unit: "ms",
-    });
-    requestCost = meter.createHistogram(LLM_METRICS.REQUEST_COST, {
-        description: "LLM request cost in USD",
-        unit: "USD",
-    });
-    tokensTotal = meter.createCounter(LLM_METRICS.TOKENS, {
-        description: "Total tokens consumed",
-    });
-    requestsTotal = meter.createCounter(LLM_METRICS.REQUESTS, {
-        description: "Total LLM requests",
-    });
-    errorsTotal = meter.createCounter(LLM_METRICS.ERRORS, {
-        description: "Total failed LLM requests",
-    });
-    initialized = true;
-}
-export function recordRequestDuration(ms, provider, model) {
-    requestDuration.record(ms, { provider, model });
-}
-export function recordRequestCost(usd, provider, model) {
-    requestCost.record(usd, { provider, model });
-}
-export function recordTokens(count, provider, model) {
-    tokensTotal.add(count, { provider, model });
-}
-export function recordRequest(provider, model) {
-    requestsTotal.add(1, { provider, model });
-}
-export function recordError(provider, model) {
-    errorsTotal.add(1, { provider, model });
-}

package/dist/spans.js

@@ -1,61 +0,0 @@
-import { trace, SpanStatusCode } from "@opentelemetry/api";
-import { LLM_ATTRS, INSTRUMENTATION_NAME } from "./types.js";
-import { recordRequestDuration, recordRequestCost, recordTokens, recordRequest, recordError, } from "./metrics.js";
-import { getConfig } from "./tracer.js";
-const tracer = trace.getTracer(INSTRUMENTATION_NAME);
-function shouldRecordContent() {
-    return getConfig()?.recordContent !== false;
-}
-function setBaseAttributes(span, input) {
-    span.setAttributes({
-        [LLM_ATTRS.PROVIDER]: input.provider,
-        [LLM_ATTRS.MODEL]: input.model,
-        [LLM_ATTRS.TEMPERATURE]: input.temperature ?? 1.0,
-        ...(shouldRecordContent() && { [LLM_ATTRS.PROMPT]: input.prompt }),
-    });
-}
-export async function traceLLMCall(input, fn) {
-    return tracer.startActiveSpan(`llm.${input.provider}.${input.model}`, async (span) => {
-        const start = performance.now();
-        setBaseAttributes(span, input);
-        try {
-            const output = await fn();
-            const duration = performance.now() - start;
-            span.setAttributes({
-                ...(shouldRecordContent() && {
-                    [LLM_ATTRS.COMPLETION]: output.completion,
-                }),
-                [LLM_ATTRS.INPUT_TOKENS]: output.inputTokens,
-                [LLM_ATTRS.OUTPUT_TOKENS]: output.outputTokens,
-                [LLM_ATTRS.COST]: output.cost,
-                [LLM_ATTRS.STATUS]: "success",
-            });
-            span.setStatus({ code: SpanStatusCode.OK });
-            recordRequest(input.provider, input.model);
-            recordRequestDuration(duration, input.provider, input.model);
-            recordRequestCost(output.cost, input.provider, input.model);
-            recordTokens(output.inputTokens + output.outputTokens, input.provider, input.model);
-            return output;
-        }
-        catch (error) {
-            const duration = performance.now() - start;
-            const message = error instanceof Error ? error.message : String(error);
-            span.setAttributes({
-                [LLM_ATTRS.COMPLETION]: "",
-                [LLM_ATTRS.INPUT_TOKENS]: 0,
-                [LLM_ATTRS.OUTPUT_TOKENS]: 0,
-                [LLM_ATTRS.COST]: 0,
-                [LLM_ATTRS.STATUS]: "error",
-                [LLM_ATTRS.ERROR]: message,
-            });
-            span.setStatus({ code: SpanStatusCode.ERROR, message });
-            recordRequest(input.provider, input.model);
-            recordRequestDuration(duration, input.provider, input.model);
-            recordError(input.provider, input.model);
-            throw error;
-        }
-        finally {
-            span.end();
-        }
-    });
-}