feat: scope model mappings by endpoint

apps/aether-gateway/src/scheduler/candidate/tests/affinity.rs

@@ -105,6 +105,7 @@ async fn same_priority_candidates_are_distributed_by_affinity_key() {
         name: "gpt-4.1-a".to_string(),
         priority: 1,
         api_formats: Some(vec!["openai:chat".to_string()]),
+        endpoint_ids: None,
     }]);
 
     let mut second = sample_row();
@@ -121,6 +122,7 @@ async fn same_priority_candidates_are_distributed_by_affinity_key() {
         name: "gpt-4.1-b".to_string(),
         priority: 1,
         api_formats: Some(vec!["openai:chat".to_string()]),
+        endpoint_ids: None,
     }]);
 
     let candidates = Arc::new(InMemoryMinimalCandidateSelectionReadRepository::seed(vec![

apps/aether-gateway/src/scheduler/candidate/tests/model.rs

@@ -36,6 +36,32 @@ fn candidate_model_names_keep_base_and_scoped_mappings() {
     assert!(!names.contains("gpt-4.1-responses"));
 }
 
+#[test]
+fn provider_model_mapping_respects_endpoint_scope() {
+    let mut row = sample_row();
+    row.endpoint_id = "endpoint-selected".to_string();
+    row.model_provider_model_mappings = Some(vec![StoredProviderModelMapping {
+        name: "gpt-4.1-endpoint".to_string(),
+        priority: 1,
+        api_formats: Some(vec!["openai:chat".to_string()]),
+        endpoint_ids: Some(vec!["endpoint-selected".to_string()]),
+    }]);
+
+    assert_eq!(
+        select_provider_model_name(&row, "openai:chat"),
+        "gpt-4.1-endpoint"
+    );
+    assert!(candidate_model_names(&row, "openai:chat").contains("gpt-4.1-endpoint"));
+
+    row.endpoint_id = "endpoint-other".to_string();
+
+    assert_eq!(
+        select_provider_model_name(&row, "openai:chat"),
+        "gpt-4.1-upstream"
+    );
+    assert!(!candidate_model_names(&row, "openai:chat").contains("gpt-4.1-endpoint"));
+}
+
 #[test]
 fn resolves_mapping_matched_model_from_key_allowed_models() {
     let mut row = sample_row();
@@ -74,6 +100,7 @@ fn resolves_requested_global_model_from_provider_model_alias() {
         name: "gpt-5.2".to_string(),
         priority: 1,
         api_formats: Some(vec!["openai:chat".to_string()]),
+        endpoint_ids: None,
     }]);
 
     let resolved = resolve_requested_global_model_name(&[row], "gpt-5.2", "openai:chat");
@@ -127,6 +154,7 @@ async fn enumerate_minimal_candidate_selection_resolves_provider_model_alias() {
         name: "gpt-5.2".to_string(),
         priority: 1,
         api_formats: Some(vec!["openai:chat".to_string()]),
+        endpoint_ids: None,
     }]);
 
     let candidates = Arc::new(InMemoryMinimalCandidateSelectionReadRepository::seed(vec![
@@ -152,6 +180,47 @@ async fn enumerate_minimal_candidate_selection_resolves_provider_model_alias() {
     assert_eq!(selection[0].selected_provider_model_name, "gpt-5.2");
 }
 
+#[tokio::test]
+async fn enumerate_minimal_candidate_selection_filters_endpoint_scoped_alias_rows() {
+    let mut selected = sample_row();
+    selected.endpoint_id = "endpoint-selected".to_string();
+    selected.key_id = "key-selected".to_string();
+    selected.global_model_name = "gpt-5".to_string();
+    selected.model_provider_model_name = "gpt-5-upstream".to_string();
+    selected.model_provider_model_mappings = Some(vec![StoredProviderModelMapping {
+        name: "gpt-5-alias".to_string(),
+        priority: 1,
+        api_formats: Some(vec!["openai:chat".to_string()]),
+        endpoint_ids: Some(vec!["endpoint-selected".to_string()]),
+    }]);
+
+    let mut other = selected.clone();
+    other.endpoint_id = "endpoint-other".to_string();
+    other.key_id = "key-other".to_string();
+
+    let candidates = Arc::new(InMemoryMinimalCandidateSelectionReadRepository::seed(vec![
+        selected, other,
+    ]));
+    let quotas = Arc::new(InMemoryProviderQuotaRepository::seed(vec![]));
+    let state = GatewayDataState::with_candidate_selection_and_quota_for_tests(candidates, quotas);
+
+    let selection = enumerate_minimal_candidate_selection_with_required_capabilities(
+        &state,
+        "openai:chat",
+        "gpt-5-alias",
+        false,
+        None,
+        None,
+        false,
+    )
+    .await
+    .expect("selection should succeed");
+
+    assert_eq!(selection.len(), 1);
+    assert_eq!(selection[0].endpoint_id, "endpoint-selected");
+    assert_eq!(selection[0].selected_provider_model_name, "gpt-5-alias");
+}
+
 #[tokio::test]
 async fn enumerate_minimal_candidate_selection_keeps_only_resolved_global_model_rows() {
     let mut exact = sample_row();
@@ -177,6 +246,7 @@ async fn enumerate_minimal_candidate_selection_keeps_only_resolved_global_model_
         name: "claude-sonnet-upstream".to_string(),
         priority: 1,
         api_formats: Some(vec!["openai:chat".to_string()]),
+        endpoint_ids: None,
     }]);
 
     let candidates = Arc::new(InMemoryMinimalCandidateSelectionReadRepository::seed(vec![
@@ -215,6 +285,7 @@ async fn enumerate_minimal_candidate_selection_allows_resolved_global_model_in_a
         name: "gpt-5-upstream".to_string(),
         priority: 1,
         api_formats: Some(vec!["openai:chat".to_string()]),
+        endpoint_ids: None,
     }]);
 
     let candidates = Arc::new(InMemoryMinimalCandidateSelectionReadRepository::seed(vec![

apps/aether-gateway/src/scheduler/candidate/tests/selection.rs

@@ -707,6 +707,7 @@ async fn selects_next_candidate_when_first_provider_quota_is_exhausted() {
         name: "gpt-4.1-primary".to_string(),
         priority: 1,
         api_formats: Some(vec!["openai:chat".to_string()]),
+        endpoint_ids: None,
     }]);
     first.key_global_priority_by_format = Some(serde_json::json!({"openai:chat": 1}));
 
@@ -721,6 +722,7 @@ async fn selects_next_candidate_when_first_provider_quota_is_exhausted() {
         name: "gpt-4.1-secondary".to_string(),
         priority: 1,
         api_formats: Some(vec!["openai:chat".to_string()]),
+        endpoint_ids: None,
     }]);
     second.key_global_priority_by_format = Some(serde_json::json!({"openai:chat": 2}));
 
@@ -776,6 +778,7 @@ async fn cooled_down_when_recent_failures_are_recorded_for_same_key() {
         name: "gpt-4.1-primary".to_string(),
         priority: 1,
         api_formats: Some(vec!["openai:chat".to_string()]),
+        endpoint_ids: None,
     }]);
     first.key_global_priority_by_format = Some(serde_json::json!({"openai:chat": 1}));
 
@@ -790,6 +793,7 @@ async fn cooled_down_when_recent_failures_are_recorded_for_same_key() {
         name: "gpt-4.1-secondary".to_string(),
         priority: 1,
         api_formats: Some(vec!["openai:chat".to_string()]),
+        endpoint_ids: None,
     }]);
     second.key_global_priority_by_format = Some(serde_json::json!({"openai:chat": 2}));
 

apps/aether-gateway/src/scheduler/candidate/tests/support.rs

@@ -39,11 +39,13 @@ pub(super) fn sample_row() -> StoredMinimalCandidateSelectionRow {
                 name: "gpt-4.1-canary".to_string(),
                 priority: 1,
                 api_formats: Some(vec!["openai:chat".to_string()]),
+                endpoint_ids: None,
             },
             StoredProviderModelMapping {
                 name: "gpt-4.1-responses".to_string(),
                 priority: 1,
                 api_formats: Some(vec!["openai:responses".to_string()]),
+                endpoint_ids: None,
             },
         ]),
         model_supports_streaming: None,