feat(api-server): add request backpressure controls

server-rs/crates/api-server/src/state.rs

@@ -27,6 +27,7 @@ use shared_contracts::creation_entry_config::CreationEntryConfigResponse;
 use shared_contracts::creative_agent::CreativeAgentSessionSnapshot;
 use spacetime_client::{SpacetimeClient, SpacetimeClientConfig, SpacetimeClientError};
 use time::OffsetDateTime;
+use tokio::sync::Semaphore;
 use tracing::{info, warn};
 
 use crate::config::AppConfig;
@@ -35,12 +36,15 @@ use crate::wechat_provider::build_wechat_provider;
 
 const ADMIN_ROLE: &str = "admin";
 
+pub type HttpRequestPermitPool = Semaphore;
+
 // 当前阶段先保留最小共享状态壳，后续逐步接入配置、客户端与平台适配。
 #[derive(Clone, Debug)]
 pub struct AppState {
     // 配置会在后续中间件、路由和平台适配接入时逐步消费。
     #[allow(dead_code)]
     pub config: AppConfig,
+    http_request_permit_pool: Option<Arc<HttpRequestPermitPool>>,
     auth_jwt_config: JwtConfig,
     admin_runtime: Option<AdminRuntime>,
     refresh_cookie_config: RefreshCookieConfig,
@@ -192,9 +196,14 @@ impl AppState {
         });
         let llm_client = build_llm_client(&config)?;
         let creative_agent_gpt5_client = build_creative_agent_gpt5_client(&config)?;
+        let http_request_permit_pool = config
+            .max_concurrent_requests
+            .map(HttpRequestPermitPool::new)
+            .map(Arc::new);
 
         Ok(Self {
             config,
+            http_request_permit_pool,
             auth_jwt_config,
             admin_runtime,
             refresh_cookie_config,
@@ -235,6 +244,10 @@ impl AppState {
         &self.refresh_cookie_config
     }
 
+    pub fn http_request_permit_pool(&self) -> Option<Arc<HttpRequestPermitPool>> {
+        self.http_request_permit_pool.clone()
+    }
+
     pub async fn upsert_creation_entry_type_config(
         &self,
         input: module_runtime::CreationEntryTypeAdminUpsertInput,