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/wallet_runtime/access.rs

@@ -3,12 +3,13 @@ use aether_wallet::{
     WalletAccessDecision, WalletAccessFailure, WalletLimitMode, WalletSnapshot, WalletStatus,
 };
 
-use crate::gateway::gateway_data::StoredGatewayAuthApiKeySnapshot;
-use crate::gateway::{AppState, GatewayError, GatewayLocalAuthRejection};
+use crate::control::GatewayLocalAuthRejection;
+use crate::data::auth::GatewayAuthApiKeySnapshot;
+use crate::{AppState, GatewayError};
 
 pub(crate) async fn resolve_wallet_auth_gate(
     state: &AppState,
-    auth_snapshot: &StoredGatewayAuthApiKeySnapshot,
+    auth_snapshot: &GatewayAuthApiKeySnapshot,
 ) -> Result<Option<WalletAccessDecision>, GatewayError> {
     if !state.has_wallet_data_reader() {
         return Ok(None);
@@ -65,7 +66,7 @@ mod tests {
     use aether_wallet::{WalletAccessFailure, WalletLimitMode, WalletSnapshot, WalletStatus};
 
     use super::{local_rejection_from_wallet_access, map_wallet_snapshot};
-    use crate::gateway::GatewayLocalAuthRejection;
+    use crate::control::GatewayLocalAuthRejection;
 
     #[test]
     fn maps_wallet_snapshot_and_derives_balance_denied() {

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

@@ -1,7 +1,5 @@
 mod access;
 mod quota;
-mod runtime;
 
 pub(crate) use access::{local_rejection_from_wallet_access, resolve_wallet_auth_gate};
 pub(crate) use quota::spawn_provider_quota_reset_worker;
-pub(crate) use runtime::settle_usage_if_needed;

apps/aether-gateway/src/wallet_runtime/quota.rs

@@ -3,7 +3,7 @@ use std::time::Duration;
 
 use tracing::warn;
 
-use crate::gateway::gateway_data::GatewayDataState;
+use crate::data::GatewayDataState;
 
 const QUOTA_RESET_INTERVAL: Duration = Duration::from_secs(60 * 60);
 
@@ -50,7 +50,7 @@ mod tests {
     };
 
     use super::{reset_due_provider_quotas_once, spawn_provider_quota_reset_worker};
-    use crate::gateway::gateway_data::GatewayDataState;
+    use crate::data::GatewayDataState;
 
     #[tokio::test]
     async fn resets_due_provider_quotas_from_runtime() {

apps/aether-gateway/src/wallet_runtime/runtime.rs

@@ -1,44 +0,0 @@
-use aether_data::repository::usage::StoredRequestUsageAudit;
-use aether_data::repository::wallet::UsageSettlementInput;
-use aether_data::{DataLayerError, DataLayerError::InvalidInput};
-
-use crate::gateway::gateway_data::GatewayDataState;
-
-pub(crate) async fn settle_usage_if_needed(
-    data: &GatewayDataState,
-    usage: &StoredRequestUsageAudit,
-) -> Result<(), DataLayerError> {
-    if !data.has_wallet_writer() || usage.billing_status != "pending" {
-        return Ok(());
-    }
-    if !matches!(usage.status.as_str(), "completed" | "failed" | "cancelled") {
-        return Ok(());
-    }
-
-    let finalized_at_unix_secs = usage
-        .finalized_at_unix_secs
-        .or(Some(usage.updated_at_unix_secs));
-    let input = UsageSettlementInput {
-        request_id: usage.request_id.clone(),
-        user_id: usage.user_id.clone(),
-        api_key_id: usage.api_key_id.clone(),
-        provider_id: usage.provider_id.clone(),
-        status: usage.status.clone(),
-        billing_status: usage.billing_status.clone(),
-        total_cost_usd: finite_cost(usage.total_cost_usd)?,
-        actual_total_cost_usd: finite_cost(usage.actual_total_cost_usd)?,
-        finalized_at_unix_secs,
-    };
-    let _ = data.settle_usage(input).await?;
-    Ok(())
-}
-
-fn finite_cost(value: f64) -> Result<f64, DataLayerError> {
-    if value.is_finite() {
-        Ok(value)
-    } else {
-        Err(InvalidInput(
-            "wallet settlement cost must be finite".to_string(),
-        ))
-    }
-}