feat: initial release v0.3.0

tests/unit/model-gateway/openai-compat-responses.test.ts

@@ -0,0 +1,67 @@
+import { beforeEach, describe, expect, it, vi } from 'vitest'
+
+const resolveOpenAICompatClientConfigMock = vi.hoisted(() =>
+  vi.fn(async () => ({
+    providerId: 'openai-compatible:node-1',
+    baseUrl: 'https://compat.example.com/v1',
+    apiKey: 'sk-test',
+  })),
+)
+
+vi.mock('@/lib/model-gateway/openai-compat/common', () => ({
+  resolveOpenAICompatClientConfig: resolveOpenAICompatClientConfigMock,
+}))
+
+import { runOpenAICompatResponsesCompletion } from '@/lib/model-gateway/openai-compat/responses'
+
+describe('model-gateway openai-compat responses executor', () => {
+  beforeEach(() => {
+    vi.clearAllMocks()
+  })
+
+  it('converts responses payload to normalized chat completion', async () => {
+    const fetchMock = vi.fn(async () => new Response(JSON.stringify({
+      output: [
+        { type: 'reasoning', text: 'think-' },
+        { type: 'output_text', text: 'hello' },
+      ],
+      usage: {
+        input_tokens: 12,
+        output_tokens: 7,
+      },
+    }), { status: 200 }))
+    vi.stubGlobal('fetch', fetchMock)
+
+    const completion = await runOpenAICompatResponsesCompletion({
+      userId: 'user-1',
+      providerId: 'openai-compatible:node-1',
+      modelId: 'gpt-4.1-mini',
+      messages: [{ role: 'user', content: 'hello' }],
+      temperature: 0.2,
+    })
+
+    expect(completion.choices[0]?.message?.content).toEqual([
+      { type: 'reasoning', text: 'think-' },
+      { type: 'text', text: 'hello' },
+    ])
+    expect(completion.usage?.prompt_tokens).toBe(12)
+    expect(completion.usage?.completion_tokens).toBe(7)
+    const firstCall = fetchMock.mock.calls[0] as unknown[] | undefined
+    expect(String(firstCall?.[0])).toBe('https://compat.example.com/v1/responses')
+  })
+
+  it('throws status-bearing error when responses endpoint fails', async () => {
+    const fetchMock = vi.fn(async () => new Response('not supported', { status: 404 }))
+    vi.stubGlobal('fetch', fetchMock)
+
+    await expect(
+      runOpenAICompatResponsesCompletion({
+        userId: 'user-1',
+        providerId: 'openai-compatible:node-1',
+        modelId: 'gpt-4.1-mini',
+        messages: [{ role: 'user', content: 'hello' }],
+        temperature: 0.2,
+      }),
+    ).rejects.toThrow('OPENAI_COMPAT_RESPONSES_FAILED: 404')
+  })
+})