AI error states are handled with clear user messaging

AI API calls fail in distinct ways: rate limits (429), network drops, content policy blocks, and model-level errors. When these surface as a blank message, a raw JSON dump, or a generic toast with no retry path, users blame the product — not the underlying API. Per iso-25010:2011 reliability.fault-tolerance, the system must degrade gracefully. A rate-limit error that tells the user to wait 60 seconds retains trust; a silent failure causes churn. Distinguishing at least two error types is the minimum required to make error messaging actionable rather than decorative.

Replace catch-all error handling with a typed error classifier that produces distinct messages for rate limits, network failures, and content policy blocks. Render the message inline in the conversation thread, not as a dismissible toast.

const getErrorMessage = (error: Error): string => {
  if (error.message.includes('429') || error.message.includes('rate limit'))
    return "Rate limit reached — wait a moment and try again."
  if (error.message.includes('content_policy'))
    return "Response blocked by content policy. Try rephrasing."
  if (error.message.includes('fetch') || error.message.includes('network'))
    return "Connection issue — check your internet and retry."
  return "Generation failed. Try again or regenerate."
}

{error && (
  <div className="flex gap-2 p-3 rounded-lg bg-destructive/10 text-destructive text-sm">
    <AlertCircleIcon className="w-4 h-4 mt-0.5 shrink-0" />
    <span>{getErrorMessage(error)} <button onClick={reload} className="underline">Retry</button></span>
  </div>
)}