feat(kiro): enhance request translation and fix streaming issues

English:
- Fix <thinking> tag parsing: only parse at response start, avoid misinterpreting discussion text
- Add token counting support using tiktoken for local estimation
- Support top_p parameter in inference config
- Handle max_tokens=-1 as maximum (32000 tokens)
- Add tool_choice and response_format parameter handling via system prompt hints
- Support multiple thinking mode detection formats (Claude API, OpenAI reasoning_effort, AMP/Cursor)
- Shorten MCP tool names exceeding 64 characters
- Fix duplicate [DONE] marker in OpenAI SSE streaming
- Enhance token usage statistics with multiple event format support
- Add code fence markers to constants

中文:
- 修复 <thinking> 标签解析：仅在响应开头解析，避免误解析讨论文本中的标签
- 使用 tiktoken 实现本地 token 计数功能
- 支持 top_p 推理配置参数
- 处理 max_tokens=-1 转换为最大值（32000 tokens）
- 通过系统提示词注入实现 tool_choice 和 response_format 参数支持
- 支持多种思考模式检测格式（Claude API、OpenAI reasoning_effort、AMP/Cursor）
- 截断超过64字符的 MCP 工具名称
- 修复 OpenAI SSE 流中重复的 [DONE] 标记
- 增强 token 使用量统计，支持多种事件格式
- 添加代码围栏标记常量

internal/translator/kiro/openai/kiro_openai_stream.go

@@ -5,7 +5,6 @@ package openai
 
 import (
 	"encoding/json"
-	"fmt"
 	"time"
 
 	"github.com/google/uuid"
@@ -34,9 +33,12 @@ func NewOpenAIStreamState(model string) *OpenAIStreamState {
 	}
 }
 
-// FormatSSEEvent formats a JSON payload as an SSE event
+// FormatSSEEvent formats a JSON payload for SSE streaming.
+// Note: This returns raw JSON data without "data:" prefix.
+// The SSE "data:" prefix is added by the Handler layer (e.g., openai_handlers.go)
+// to maintain architectural consistency and avoid double-prefix issues.
 func FormatSSEEvent(data []byte) string {
-	return fmt.Sprintf("data: %s", string(data))
+	return string(data)
 }
 
 // BuildOpenAISSETextDelta creates an SSE event for text content delta
@@ -130,9 +132,12 @@ func BuildOpenAISSEUsage(state *OpenAIStreamState, usageInfo usage.Detail) strin
 	return FormatSSEEvent(result)
 }
 
-// BuildOpenAISSEDone creates the final [DONE] SSE event
+// BuildOpenAISSEDone creates the final [DONE] SSE event.
+// Note: This returns raw "[DONE]" without "data:" prefix.
+// The SSE "data:" prefix is added by the Handler layer (e.g., openai_handlers.go)
+// to maintain architectural consistency and avoid double-prefix issues.
 func BuildOpenAISSEDone() string {
-	return "data: [DONE]"
+	return "[DONE]"
 }
 
 // buildBaseChunk creates a base chunk structure for streaming