feat(ai-pipeline): 支持 reasoning signature 及媒体内容块的跨格式流式转换

- 在 CanonicalStreamEvent 中新增 ReasoningSignature 和 ContentPart 变体，CanonicalContentPart 枚举覆盖图片/文件/音频
- Gemini 流解析器提取 thoughtSignature，并对 inlineData/fileData 等非文本 part 生成 ContentPart 事件
- Claude 流聚合支持 thinking_delta / signature_delta，输出 thinking block 携带 signature 字段
- Gemini 同步响应聚合重写，支持媒体 part 和 reasoning signature 的完整还原
- 跨格式矩阵（gemini↔claude↔openai）补全图片块双向转换及 reasoning signature 传递
- 请求转换层（to/from openai_chat）补全 Claude/Gemini 的 thinking、图片、工具调用字段映射
- 新增/扩展测试：inline image 双向重写、thinking signature 聚合、跨格式 sync product 媒体字段

apps/aether-gateway/src/tests/usage/local.rs

@@ -36,7 +36,7 @@ where
     let mut stored = None;
     // Usage terminal events are written on a shared background runtime; under full-suite parallel
     // load they can lag noticeably behind the request/response assertion path.
-    let timeout = std::time::Duration::from_secs(30);
+    let timeout = std::time::Duration::from_secs(60);
     let deadline = tokio::time::Instant::now() + timeout;
     loop {
         stored = repository