prompt() implement all specified arguments

[gr2m]

As of September 1st, we document the following usage example for prompt():
(see the latest spec at https://github.com/copilot-extensions/preview-sdk.js#promptmessage-options)

await prompt({
  model: "gpt-4o",
  token: process.env.TOKEN,
  system: "You are a helpful assistant.",
  messages: [
    { role: "user", content: "What is the capital of France?" },
    { role: "assistant", content: "The capital of France is Paris." },
    {
      role: "user",
      content: [
        [
          { type: "text", text: "What about this country?" },
          {
            type: "image_url",
            image_url: urlToImageOfFlagOfSpain,
          },
        ],
      ],
    },
  ],
  // GitHub recommends using your GitHub username, the name of your application
  // https://docs.github.com/en/rest/using-the-rest-api/getting-started-with-the-rest-api?apiVersion=2022-11-28#user-agent
  userAgent: "gr2m/my-app v1.2.3",
  // set an alternative chat completions endpoint
  endpoint: "https://models.inference.ai.azure.com/chat/completions",
  // compare https://platform.openai.com/docs/guides/function-calling/configuring-function-calling-behavior-using-the-tool_choice-parameter
  toolChoice: "auto",
  tools: [
    {
      type: "function",
      function: {
        name: "get_weather",
        strict: true,
        parameters: {
          type: "object",
          properties: {
            location: { type: "string" },
            unit: { type: "string", enum: ["c", "f"] },
          },
          required: ["location", "unit"],
          additionalProperties: False,
        },
      },
    },
  ],
  // configuration related to the request transport layer
  request: {
    // for mocking, proxying, client certificates, etc.
    fetch: myCustomFetch,
    // hook into request life cycle for complex authentication strategies, retries, throttling, etc
    // compare options.request.hook from https://github.com/octokit/request.js
    hook: myCustomHook,
    // Use an `AbortController` instance to cancel a request
    signal: myAbortController.signal,
  },
});

Out of the documented arguments, only model, token, and request.fetch is implemented as of September 1st.