flynn/src/models/openai.test.ts

import { describe, it, expect, vi } from 'vitest';
import { OpenAIClient } from './openai.js';

const { mockCreate, mockOpenAIConstructor } = vi.hoisted(() => {
  const mockCreate = vi.fn().mockResolvedValue({
    choices: [{ message: { content: 'Hello from GPT!' }, finish_reason: 'stop' }],
    usage: { prompt_tokens: 10, completion_tokens: 5 },
  });
  const mockOpenAIConstructor = vi.fn().mockImplementation(() => ({
    chat: {
      completions: {
        create: mockCreate,
      },
    },
  }));
  return { mockCreate, mockOpenAIConstructor };
});

vi.mock('openai', () => ({
  default: mockOpenAIConstructor,
}));

describe('OpenAIClient', () => {
  it('sets request timeout and disables SDK retries', () => {
    new OpenAIClient({
      apiKey: 'test-key',
      model: 'gpt-4o',
    });

    expect(mockOpenAIConstructor).toHaveBeenCalledWith(expect.objectContaining({
      timeout: 20_000,
      maxRetries: 0,
    }));
  });

  it('sends messages and returns response', async () => {
    const client = new OpenAIClient({
      apiKey: 'test-key',
      model: 'gpt-4o',
    });

    const response = await client.chat({
      messages: [{ role: 'user', content: 'Hello' }],
    });

    expect(response.content).toBe('Hello from GPT!');
    expect(response.stopReason).toBe('end_turn');
    expect(response.usage.inputTokens).toBe(10);
    expect(response.usage.outputTokens).toBe(5);
  });
});

describe('OpenAIClient tool use', () => {
  it('passes tools to API and parses tool_calls response', async () => {
    mockCreate.mockResolvedValueOnce({
      choices: [{
        message: {
          content: null,
          tool_calls: [{
            id: 'call_1',
            type: 'function',
            function: { name: 'shell.exec', arguments: '{"command":"ls"}' },
          }],
        },
        finish_reason: 'tool_calls',
      }],
      usage: { prompt_tokens: 20, completion_tokens: 15 },
    });

    const client = new OpenAIClient({
      apiKey: 'test-key',
      model: 'gpt-4o',
    });

    const response = await client.chat({
      messages: [{ role: 'user', content: 'list files' }],
      tools: [{
        name: 'shell.exec',
        description: 'Run shell command',
        input_schema: { type: 'object', properties: { command: { type: 'string' } }, required: ['command'] },
      }],
    });

    expect(response.stopReason).toBe('tool_use');
    expect(response.toolCalls).toHaveLength(1);
    expect(response.toolCalls![0]).toEqual({
      id: 'call_1',
      name: 'shell.exec',
      args: { command: 'ls' },
    });
  });

  it('maps finish_reason "tool_calls" with empty tool_calls to end_turn', async () => {
    mockCreate.mockResolvedValueOnce({
      choices: [{
        message: {
          content: 'I tried to call a tool but none matched.',
          tool_calls: [],
        },
        finish_reason: 'tool_calls',
      }],
      usage: { prompt_tokens: 15, completion_tokens: 10 },
    });

    const client = new OpenAIClient({
      apiKey: 'test-key',
      model: 'gpt-4o',
    });

    const response = await client.chat({
      messages: [{ role: 'user', content: 'do something' }],
    });

    expect(response.stopReason).toBe('end_turn');
    expect(response.toolCalls).toBeUndefined();
  });

  it('maps finish_reason "length" to max_tokens', async () => {
    mockCreate.mockResolvedValueOnce({
      choices: [{
        message: { content: 'Truncated output...' },
        finish_reason: 'length',
      }],
      usage: { prompt_tokens: 100, completion_tokens: 4096 },
    });

    const client = new OpenAIClient({
      apiKey: 'test-key',
      model: 'gpt-4o',
    });

    const response = await client.chat({
      messages: [{ role: 'user', content: 'write a long essay' }],
    });

    expect(response.stopReason).toBe('max_tokens');
  });
});