fix: preserve upstream stream accept negotiation

apps/aether-gateway/src/ai_serving/planner/common.rs

@@ -1,8 +1,11 @@
 use axum::body::Bytes;
 
 use crate::ai_serving::{
+    endpoint_config_forces_upstream_stream_policy as endpoint_config_forces_upstream_stream_policy_impl,
+    enforce_request_body_stream_field as enforce_request_body_stream_field_impl,
     force_upstream_streaming_for_provider as force_upstream_streaming_for_provider_impl,
     is_json_request, parse_direct_request_body as parse_direct_request_body_impl,
+    resolve_upstream_is_stream_from_endpoint_config as resolve_upstream_is_stream_from_endpoint_config_impl,
 };
 pub(crate) use crate::ai_serving::{
     CLAUDE_CHAT_STREAM_PLAN_KIND, CLAUDE_CHAT_SYNC_PLAN_KIND, CLAUDE_CLI_STREAM_PLAN_KIND,
@@ -46,7 +49,7 @@ pub(crate) fn resolve_upstream_is_stream_for_provider(
 ) -> bool {
     let hard_requires_streaming = hard_requires_streaming
         || force_upstream_streaming_for_provider(provider_type, provider_api_format);
-    aether_ai_formats::resolve_upstream_is_stream_from_endpoint_config(
+    resolve_upstream_is_stream_from_endpoint_config_impl(
         endpoint_config,
         client_is_stream,
         hard_requires_streaming,
@@ -56,7 +59,7 @@ pub(crate) fn resolve_upstream_is_stream_for_provider(
 pub(crate) fn endpoint_config_forces_body_stream_field(
     endpoint_config: Option<&serde_json::Value>,
 ) -> bool {
-    aether_ai_formats::endpoint_config_forces_upstream_stream_policy(endpoint_config)
+    endpoint_config_forces_upstream_stream_policy_impl(endpoint_config)
 }
 
 pub(crate) fn request_requires_body_stream_field(
@@ -75,7 +78,7 @@ pub(crate) fn enforce_provider_body_stream_policy(
     upstream_is_stream: bool,
     require_body_stream_field: bool,
 ) {
-    aether_ai_formats::enforce_request_body_stream_field(
+    enforce_request_body_stream_field_impl(
         provider_request_body,
         provider_api_format,
         upstream_is_stream,

apps/aether-gateway/src/ai_serving/planner/standard/openai/plan_builders/stream.rs

@@ -18,6 +18,18 @@ use crate::ai_serving::transport::{
 };
 use crate::{AiExecutionDecision, GatewayError};
 
+fn effective_stream_accept_mode(
+    payload_upstream_is_stream: bool,
+    provider_request_body: &serde_json::Value,
+) -> bool {
+    payload_upstream_is_stream
+        || provider_request_body
+            .as_object()
+            .and_then(|body| body.get("stream"))
+            .and_then(serde_json::Value::as_bool)
+            .unwrap_or(false)
+}
+
 pub(crate) fn build_openai_chat_stream_plan_from_decision(
     parts: &http::request::Parts,
     body_json: &serde_json::Value,
@@ -80,6 +92,8 @@ pub(crate) fn build_openai_chat_stream_plan_from_decision(
         }
         provider_request_body
     };
+    let effective_upstream_is_stream =
+        effective_stream_accept_mode(payload.upstream_is_stream, &provider_request_body_value);
     let extra_headers = std::mem::take(&mut payload.extra_headers);
     let mut provider_request_headers =
         build_standard_plan_fallback_headers(StandardPlanFallbackHeadersInput {
@@ -93,9 +107,9 @@ pub(crate) fn build_openai_chat_stream_plan_from_decision(
             content_type: payload.content_type.as_deref(),
             provider_api_format: core.provider_api_format.as_str(),
             client_api_format: core.client_api_format.as_str(),
-            upstream_is_stream: payload.upstream_is_stream,
+            upstream_is_stream: effective_upstream_is_stream,
             build_from_request_when_empty: true,
-            accept_policy: StandardPlanFallbackAcceptPolicy::TextEventStreamIfStreaming,
+            accept_policy: StandardPlanFallbackAcceptPolicy::TextEventStreamIfStreamingOrWildcard,
         });
     let content_type = payload
         .content_type
@@ -168,14 +182,18 @@ pub(crate) fn build_openai_responses_stream_plan_from_decision(
         .as_ref()
         .and_then(|context| context.get("envelope_name"))
         .and_then(serde_json::Value::as_str);
-    let accept_policy = if payload.upstream_is_stream
+    let effective_upstream_is_stream =
+        effective_stream_accept_mode(payload.upstream_is_stream, &provider_request_body_value);
+    let accept_policy = if effective_upstream_is_stream
         && provider_adaptation_requires_eventstream_accept(
             envelope_name,
             core.provider_api_format.as_str(),
         ) {
         StandardPlanFallbackAcceptPolicy::ProviderEventStreamIfMissing
-    } else {
+    } else if envelope_name.is_some() {
         StandardPlanFallbackAcceptPolicy::TextEventStreamIfStreaming
+    } else {
+        StandardPlanFallbackAcceptPolicy::TextEventStreamIfStreamingOrWildcard
     };
     let mut provider_request_headers =
         build_standard_plan_fallback_headers(StandardPlanFallbackHeadersInput {
@@ -189,7 +207,7 @@ pub(crate) fn build_openai_responses_stream_plan_from_decision(
             content_type: payload.content_type.as_deref(),
             provider_api_format: core.provider_api_format.as_str(),
             client_api_format: core.client_api_format.as_str(),
-            upstream_is_stream: payload.upstream_is_stream,
+            upstream_is_stream: effective_upstream_is_stream,
             build_from_request_when_empty: false,
             accept_policy,
         });
@@ -231,7 +249,7 @@ pub(crate) fn build_openai_responses_stream_plan_from_decision(
         plan_url = %plan.url,
         client_api_format = %plan.client_api_format,
         provider_api_format = %plan.provider_api_format,
-        upstream_is_stream = payload.upstream_is_stream,
+        upstream_is_stream = effective_upstream_is_stream,
         compact,
         "gateway built local openai responses stream execution plan"
     );