aethershell 11.0.1

// AetherShell TUI Multimodal Demo
// Launch with: ae tui
//
// This demo showcases multimodal AI capabilities:
// - Image analysis
// - Audio processing
// - Video understanding
// - Mixed media conversations
//

// Analyze an image with GPT-4o
// This example calls a live AI provider. Without one it says so and stops,
// rather than failing partway through with a provider error.
if env("AETHER_AI") == null {
  print("Skipping: set AETHER_AI (openai, ollama, irongate or compat) plus the matching API key to run this example.")
}
if env("AETHER_AI") == null { exit(0) }

ai "openai:gpt-4o" {
  prompt: "What's in this image? Describe it in detail.",
  media: ["./demos/sample_image.jpg"]
}

// Audio transcription and analysis
ai "openai:gpt-4o" {
  prompt: "Transcribe this audio and summarize the main points.",
  media: ["./demos/sample_audio.mp3"]
}

// Video understanding
ai "openai:gpt-4o" {
  prompt: "What's happening in this video? Describe the key actions.",
  media: ["./demos/sample_video.mp4"]
}

// Multiple images comparison
ai "openai:gpt-4o" {
  prompt: "Compare these two images and describe the differences.",
  media: ["./demos/image1.jpg", "./demos/image2.jpg"]
}

// Mixed media: text, image, and context
ai "openai:gpt-4o" {
  prompt: "Based on this diagram and the audio explanation, summarize the concept.",
  media: ["./demos/diagram.png", "./demos/explanation.mp3"]
}

// The TUI provides:
// - Media preview pane (images, audio waveforms, video thumbnails)
// - Drag-and-drop media attachment
// - Media gallery navigation
// - Format support: JPG, PNG, MP3, WAV, MP4, WebM
// - Base64 encoding for API transmission
// - Progress indicators for large files
//

print "TUI Multimodal Demo Ready! Launch with: ae tui"
print "Note: Replace sample media paths with your actual files"