feat: factuality eval with assistant response

apps/studio/evals/assistant.eval.ts

@@ -0,0 +1,31 @@
+import { openai } from '@ai-sdk/openai'
+import { Factuality } from 'autoevals'
+import { Eval } from 'braintrust'
+import { generateAssistantResponse } from 'lib/ai/generate-assistant-response'
+import assert from 'node:assert'
+
+assert(process.env.BRAINTRUST_PROJECT_ID, 'BRAINTRUST_PROJECT_ID is not set')
+assert(process.env.OPENAI_API_KEY, 'OPENAI_API_KEY is not set')
+
+Eval('Assistant', {
+  projectId: process.env.BRAINTRUST_PROJECT_ID,
+  data: () => {
+    return [
+      {
+        input: 'Hello!',
+        expected: 'Hello, world!',
+      },
+    ]
+  },
+  task: async (input) => {
+    const result = await generateAssistantResponse({
+      model: openai('gpt-5'),
+      messages: [{ id: '1', role: 'user', parts: [{ type: 'text', text: input }] }],
+      tools: {},
+    })
+
+    const text = await result.text
+    return text
+  },
+  scores: [Factuality],
+})