feat: add audio input support to v2 providers

Add Audio struct alongside Image for sending audio attachments to
multimodal LLMs. OpenAI uses input_audio content parts (wav/mp3),
Google Gemini uses genai.NewPartFromBytes, and Anthropic skips
audio gracefully since it's not supported.

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

v2/provider/provider.go

@@ -8,6 +8,7 @@ type Message struct {
 	Role       string
 	Content    string
 	Images     []Image
+	Audio      []Audio
 	ToolCalls  []ToolCall
 	ToolCallID string
 }
@@ -19,6 +20,13 @@ type Image struct {
 	ContentType string
 }
 
+// Audio represents an audio attachment at the provider level.
+type Audio struct {
+	URL         string
+	Base64      string
+	ContentType string
+}
+
 // ToolCall represents a tool invocation requested by the model.
 type ToolCall struct {
 	ID        string