Add multimodal content and tool calls support

Author: dev2k6

internal/api/openai.go

@@ -3,9 +3,42 @@ package api
 // OpenAI-compatible types (client-facing)
 
 type OpenAIMessage struct {
-	Role    string `json:"role"`
-	Content string `json:"content"`
-	Name    string `json:"name,omitempty"`
+	Role          string        `json:"role"`
+	Content       interface{}   `json:"content,omitempty"`
+	Name          string        `json:"name,omitempty"`
+	ToolCalls     []ToolCall    `json:"tool_calls,omitempty"`
+	ToolCallID    string        `json:"tool_call_id,omitempty"`
+	Refusal       string        `json:"refusal,omitempty"`
+	Audio         *MessageAudio `json:"audio,omitempty"`
+}
+
+type ContentPart struct {
+	Type     string    `json:"type"`
+	Text     string    `json:"text,omitempty"`
+	ImageURL *ImageURL `json:"image_url,omitempty"`
+}
+
+type ImageURL struct {
+	URL       string `json:"url"`
+	Detail    string `json:"detail,omitempty"`
+	Modalities string `json:"modalities,omitempty"`
+}
+
+type ToolCall struct {
+	ID       string      `json:"id"`
+	Type     string      `json:"type"`
+	Function FunctionCall `json:"function"`
+}
+
+type FunctionCall struct {
+	Name      string `json:"name"`
+	Arguments string `json:"arguments"`
+}
+
+type MessageAudio struct {
+	ID       string `json:"id"`
+	Data     string `json:"data"`
+	Duration float64 `json:"duration"`
 }
 
 type OpenAIChatRequest struct {

internal/proxy/convert.go

@@ -10,16 +10,50 @@ import (
 func ConvertMessages(openAIMsgs []api.OpenAIMessage) []api.CCMessage {
 	var ccMsgs []api.CCMessage
 	for _, m := range openAIMsgs {
+		contentParts := parseContent(m.Content)
 		ccMsgs = append(ccMsgs, api.CCMessage{
-			Role: m.Role,
-			Content: []api.CCContentPart{
-				{Type: "text", Text: m.Content},
-			},
+			Role:    m.Role,
+			Content: contentParts,
 		})
 	}
 	return ccMsgs
 }
 
+func parseContent(content interface{}) []api.CCContentPart {
+	switch v := content.(type) {
+	case string:
+		if v == "" {
+			return nil
+		}
+		return []api.CCContentPart{{Type: "text", Text: v}}
+	case []any:
+		var parts []api.CCContentPart
+		for _, part := range v {
+			if partMap, ok := part.(map[string]any); ok {
+				if typ, ok := partMap["type"].(string); ok {
+					p := api.CCContentPart{Type: typ}
+					if text, ok := partMap["text"].(string); ok {
+						p.Text = text
+					}
+					// Note: CommandCode API may not support image_url
+					// For now, we skip image parts
+					if imgURL, ok := partMap["image_url"].(map[string]any); ok {
+						if url, ok := imgURL["url"].(string); ok {
+							// Try to extract text from image (e.g., base64 data)
+							// or just include the URL as text
+							p.Text = p.Text + "\n[Image URL: " + url + "]"
+						}
+					}
+					parts = append(parts, p)
+				}
+			}
+		}
+		return parts
+	default:
+		return nil
+	}
+}
+
 // Extract system message and remaining messages
 func ExtractSystem(msgs []api.OpenAIMessage) (string, []api.OpenAIMessage) {
 	var system strings.Builder
@@ -29,7 +63,9 @@ func ExtractSystem(msgs []api.OpenAIMessage) (string, []api.OpenAIMessage) {
 			if system.Len() > 0 {
 				system.WriteString("\n")
 			}
-			system.WriteString(m.Content)
+			if contentStr, ok := m.Content.(string); ok {
+				system.WriteString(contentStr)
+			}
 		} else {
 			rest = append(rest, m)
 		}

internal/proxy/proxy.go

@@ -73,7 +73,7 @@ func (p *Proxy) BuildRequest(openAIReq api.OpenAIChatRequest) (api.CCRequestBody
 			System:      system,
 			MaxTokens:   maxTokens,
 			Temperature: temperature,
-			Stream:      true,
+			Stream:      openAIReq.Stream,
 		},
 		ThreadID: uuid.New().String(),
 	}