feat: experimental span filter sampler

bun.lock

@@ -1,10 +1,12 @@
 {
   "lockfileVersion": 1,
+  "configVersion": 0,
   "workspaces": {
     "": {
       "name": "judgeval",
       "dependencies": {
         "@opentelemetry/api": "^1.9.0",
+        "@opentelemetry/context-async-hooks": "^2.2.0",
         "@opentelemetry/core": "^2.2.0",
         "@opentelemetry/exporter-trace-otlp-http": "^0.207.0",
         "@opentelemetry/resources": "^2.2.0",

examples/simple_chat/instrumentation.ts

@@ -3,16 +3,10 @@ import { Judgeval, type NodeTracer } from "judgeval";
 
 export const client = Judgeval.create();
 
-const initPromise = client.nodeTracer
-  .create({
-    projectName: "auto_instrumentation_example",
-    enableEvaluation: true,
-    enableMonitoring: true,
-    instrumentations: [new OpenAIInstrumentation()],
-  })
-  .then((t: NodeTracer) => {
-    return t;
-  });
+const initPromise = client.nodeTracer.create({
+  projectName: "auto_instrumentation_example",
+  instrumentations: [new OpenAIInstrumentation()],
+});
 
 export async function getTracer(): Promise<NodeTracer> {
   return await initPromise;

package.json

@@ -1,6 +1,6 @@
 {
   "name": "judgeval",
-  "version": "0.7.2",
+  "version": "0.7.3",
   "description": "JavaScript/TypeScript client for Judgment evaluation platform",
   "main": "./dist/index.cjs",
   "module": "./dist/index.mjs",
@@ -26,7 +26,7 @@
     "lint:fix": "eslint src/**/*.ts --fix --no-warn-ignored",
     "format": "prettier --write \"src/**/*.ts\"",
     "generate-client": "bunx tsx scripts/generate-client.ts && bun run format",
-    "test": "bun test",
+    "test": "bun test --coverage",
     "clean": "rimraf dist",
     "prepublishOnly": "bun run clean && bun run build",
     "example": "bun scripts/run-example.ts"
@@ -48,6 +48,7 @@
   },
   "dependencies": {
     "@opentelemetry/api": "^1.9.0",
+    "@opentelemetry/context-async-hooks": "^2.2.0",
     "@opentelemetry/core": "^2.2.0",
     "@opentelemetry/exporter-trace-otlp-http": "^0.207.0",
     "@opentelemetry/resources": "^2.2.0",

src/env.ts

@@ -14,13 +14,9 @@ export const JUDGMENT_API_URL = getEnvVar(
   "JUDGMENT_API_URL",
   "https://api.judgmentlabs.ai",
 );
-export const JUDGMENT_LLM_PROXY_URL = getEnvVar(
-  "JUDGMENT_LLM_PROXY_URL",
-  "https://api.judgmentlabs.ai/llm/proxy/v1",
-);
 export const JUDGMENT_DEFAULT_GPT_MODEL = getEnvVar(
   "JUDGMENT_DEFAULT_GPT_MODEL",
-  "gpt-4.1",
+  "gpt-5-mini",
 );
 export const JUDGMENT_ENABLE_MONITORING = getEnvVar(
   "JUDGMENT_ENABLE_MONITORING",

src/scorers/promptScorer/PromptScorer.ts

@@ -1,5 +1,5 @@
-import type { ScorerConfig } from "../../internal/api/models";
 import { APIScorerType } from "../../data/APIScorerType";
+import type { ScorerConfig } from "../../internal/api/models";
 import { BaseScorer } from "../BaseScorer";
 
 export interface PromptScorerConfig {
@@ -91,9 +91,11 @@ export class PromptScorer extends BaseScorer {
     if (this._options) {
       kwargs.options = this._options;
     }
+
     if (this._model) {
       kwargs.model = this._model;
     }
+
     if (this._description) {
       kwargs.description = this._description;
     }

src/scorers/promptScorer/PromptScorerFactory.ts

@@ -5,7 +5,8 @@ import type {
   FetchPromptScorersRequest,
   FetchPromptScorersResponse,
 } from "../../internal/api/models";
-import { PromptScorer, type PromptScorerConfig } from "./PromptScorer";
+import { Logger } from "../../utils";
+import { PromptScorer } from "./PromptScorer";
 
 export class PromptScorerFactory {
   private readonly client: JudgmentApiClient;
@@ -17,12 +18,12 @@ export class PromptScorerFactory {
     this.isTrace = isTrace;
   }
 
-  async get(name: string): Promise<PromptScorer> {
+  async get(name: string): Promise<PromptScorer | null> {
     const cacheKey = this.getCacheKey(name);
     const cached = PromptScorerFactory.cache.get(cacheKey);
 
     if (cached) {
-      return this.createFromModel(cached, name);
+      return this._create(cached, name);
     }
 
     try {
@@ -51,34 +52,18 @@ export class PromptScorerFactory {
       }
 
       PromptScorerFactory.cache.set(cacheKey, scorer);
-      return this.createFromModel(scorer, name);
+      return this._create(scorer, name);
     } catch (error) {
-      if (error instanceof Error) {
-        throw error;
-      }
-      throw new Error(`Failed to fetch prompt scorer '${name}': ${error}`);
-    }
-  }
-
-  create(config: PromptScorerConfig): PromptScorer {
-    if (!config.name) {
-      throw new Error("Name is required");
-    }
-    if (!config.prompt) {
-      throw new Error("Prompt is required");
+      Logger.error(`Failed to fetch prompt scorer '${name}': ${error}`);
+      return null;
     }
-
-    return new PromptScorer({
-      ...config,
-      isTrace: this.isTrace,
-    });
   }
 
-  private createFromModel(model: APIPromptScorer, name: string): PromptScorer {
+  private _create(scorer: APIPromptScorer, name: string): PromptScorer {
     let options: Record<string, number> | undefined;
-    if (model.options && typeof model.options === "object") {
+    if (scorer.options && typeof scorer.options === "object") {
       options = {};
-      for (const [key, value] of Object.entries(model.options)) {
+      for (const [key, value] of Object.entries(scorer.options)) {
         if (typeof value === "number") {
           options[key] = value;
         }
@@ -87,11 +72,11 @@ export class PromptScorerFactory {
 
     return new PromptScorer({
       name,
-      prompt: model.prompt,
-      threshold: model.threshold,
+      prompt: scorer.prompt,
+      threshold: scorer.threshold,
       options: options ?? {},
-      model: model.model ?? JUDGMENT_DEFAULT_GPT_MODEL,
-      description: model.description ?? "",
+      model: scorer.model ?? JUDGMENT_DEFAULT_GPT_MODEL,
+      description: scorer.description ?? "",
       isTrace: this.isTrace,
     });
   }