feat: add image generation for OpenAI DALL-E and Gemini Imagen

New `/v1/images/generations` endpoint proxies DALL-E 2/3 (OpenAI)
and Imagen 3 (Gemini). Same auth/logging as chat completions.

- Add ImageGenerationRequest/Response models
- Extend Provider interface with ImageGeneration()
- OpenAI: forward to /v1/images/generations
- Gemini: call /v1beta/models/{model}:predict, map OpenAI params
- Circuit breaker wraps image gen like chat completions
- Model routing: dall-e* -> openai, imagen*/gemini* -> gemini
- Unsupported providers (deepseek/moonshot/grok/ollama) return error
- Fix pre-existing CachedContentTokenCount bug in StreamGemini

internal/providers/grok.go

@@ -2,13 +2,13 @@ package providers
 
 import (
 	"context"
-	"time"
 	"encoding/json"
 	"fmt"
+	"time"
 
+	"github.com/go-resty/resty/v2"
 	"gophergate/internal/config"
 	"gophergate/internal/models"
-	"github.com/go-resty/resty/v2"
 )
 
 type GrokProvider struct {
@@ -83,7 +83,7 @@ func (p *GrokProvider) ChatCompletionStream(ctx context.Context, req *models.Uni
 	}
 
 	ch := make(chan *models.ChatCompletionStreamResponse)
-	
+
 	go func() {
 		defer close(ch)
 		err := StreamOpenAI(resp.RawBody(), ch)
@@ -94,3 +94,7 @@ func (p *GrokProvider) ChatCompletionStream(ctx context.Context, req *models.Uni
 
 	return ch, nil
 }
+
+func (p *GrokProvider) ImageGeneration(ctx context.Context, req *models.ImageGenerationRequest) (*models.ImageGenerationResponse, error) {
+	return nil, fmt.Errorf("grok does not support image generation")
+}