refactor: 拆分 gateway 单体为独立 crate，新增 systemd 部署方案

将 gateway 内部的 model-fetch、provider-transport、scheduler-core、
usage-runtime、video-tasks-core 模块提取为独立 crate；重构 gateway
内部模块结构（state/router/cache/data/query 等）；移除大量遗留模块
文件；新增 systemd 二进制部署骨架及相关文档；更新前端 usage 相关
API 和组件。

apps/aether-gateway/src/tunnel/embedded/hub.rs

@@ -832,9 +832,7 @@ impl HubStats {
 mod tests {
     use aether_runtime::bounded_queue;
 
-    use super::{
-        protocol, ControlPlaneClient, HubRouter, ProxyConn, MAX_REQUEST_BODY_FRAME_SIZE,
-    };
+    use super::{protocol, ControlPlaneClient, HubRouter, ProxyConn, MAX_REQUEST_BODY_FRAME_SIZE};
     use axum::extract::ws::Message;
     use bytes::Bytes;
     use std::collections::HashMap;

apps/aether-gateway/src/tunnel/embedded/local_relay.rs

@@ -13,6 +13,8 @@ use bytes::BytesMut;
 use futures_util::StreamExt;
 use tracing::warn;
 
+use crate::api::response::apply_streaming_response_headers;
+
 use super::hub::{LocalBodyEvent, LocalStream};
 use super::protocol;
 use super::AppState;
@@ -261,6 +263,7 @@ pub async fn relay_request(
     let mut builder = Response::builder().status(response_head.status);
     if let Some(headers) = builder.headers_mut() {
         append_headers(headers, &response_head.headers);
+        apply_streaming_response_headers(headers);
     }
     match builder.body(Body::from_stream(body_stream)) {
         Ok(response) => maybe_hold_axum_response_permit(response, request_permit),

apps/aether-gateway/src/tunnel/mod.rs

@@ -24,7 +24,7 @@ use self::embedded::{AppState as TunnelAppState, ConnConfig, ControlPlaneClient}
 use super::api::response::{build_client_response, build_local_http_error_response};
 use super::constants::TRACE_ID_HEADER;
 use super::error::GatewayError;
-use super::gateway_data::GatewayDataState;
+use super::data::GatewayDataState;
 use super::headers::{extract_or_generate_trace_id, should_skip_request_header};
 use super::AppState;
 
@@ -801,8 +801,7 @@ fn parse_embedded_tunnel_heartbeat_request(
 mod tests {
     use super::{
         apply_embedded_tunnel_heartbeat, apply_embedded_tunnel_node_status, current_unix_secs,
-        tunnel_attachment_key, TunnelAttachmentDirectory, TunnelAttachmentRecord,
-        GatewayDataState,
+        tunnel_attachment_key, GatewayDataState, TunnelAttachmentDirectory, TunnelAttachmentRecord,
     };
     use aether_data::repository::proxy_nodes::{
         InMemoryProxyNodeRepository, ProxyNodeReadRepository, StoredProxyNode,