fix: cache GET /api/models for five minutes

CHANGELOG.md

@@ -7,8 +7,12 @@ Write each change in both `### English` and `### 中文` under `## Unreleased`.
 
 ### English
 
+- Cache `GET /api/models` for 5 minutes so the console catalog page does not re-hit WorkBuddy or Trae on every load; `?refresh=1` still fetches live. Overview stays uncached.
+
 ### 中文
 
+- `GET /api/models` 缓存 5 分钟，控制台模型页不再每次都打 WorkBuddy / Trae 目录；`?refresh=1` 仍即时拉取。Overview 不缓存。
+
 ## 0.2.48 - 2026-09-06
 
 ### English

internal/api/auth_test.go

@@ -9,6 +9,7 @@ import (
 	"net/http"
 	"net/http/httptest"
 	"strings"
+	"sync/atomic"
 	"testing"
 	"time"
 
@@ -228,6 +229,65 @@ func TestModelsAPICatalogFailureUses503(t *testing.T) {
 	}
 }
 
+type countingCatalog struct {
+	hits   atomic.Int32
+	models []providers.ModelInfo
+}
+
+func (c *countingCatalog) Models(context.Context, string) ([]providers.ModelInfo, error) {
+	c.hits.Add(1)
+	return c.models, nil
+}
+
+func TestModelsAPICachesCatalogForFiveMinutes(t *testing.T) {
+	srv := New(config.Config{
+		Host: "127.0.0.1", Port: 3010, ProxyAPIKey: "secret",
+		QoderHome: t.TempDir(), DataDir: t.TempDir(),
+	})
+	defer srv.Close()
+	catalog := &countingCatalog{models: []providers.ModelInfo{{
+		NativeModel: "glm-5.3", PublicModel: "glm-5.3", DisplayName: "GLM",
+	}}}
+	srv.pool.Upsert(accounts.Item{ID: "wb-cn", Provider: "workbuddy", Runtime: string(providers.RuntimeInProcess)})
+	srv.providers.Register(providers.Adapter{ID: "workbuddy", Models: catalog})
+
+	getModels := func(path string) *httptest.ResponseRecorder {
+		req := httptest.NewRequest(http.MethodGet, path, nil)
+		req.Header.Set("Authorization", "Bearer secret")
+		rec := httptest.NewRecorder()
+		srv.Handler().ServeHTTP(rec, req)
+		return rec
+	}
+
+	first := getModels("/api/models?account=wb-cn")
+	if first.Code != http.StatusOK {
+		t.Fatalf("first /api/models = %d %s", first.Code, first.Body.String())
+	}
+	second := getModels("/api/models?account=wb-cn")
+	if second.Code != http.StatusOK {
+		t.Fatalf("cached /api/models = %d %s", second.Code, second.Body.String())
+	}
+	if catalog.hits.Load() != 1 {
+		t.Fatalf("cached /api/models hits = %d, want 1", catalog.hits.Load())
+	}
+
+	overview := getModels("/api/overview")
+	if overview.Code != http.StatusOK {
+		t.Fatalf("/api/overview = %d %s", overview.Code, overview.Body.String())
+	}
+	if catalog.hits.Load() != 2 {
+		t.Fatalf("overview must not use /api/models cache, hits = %d", catalog.hits.Load())
+	}
+
+	refreshed := getModels("/api/models?account=wb-cn&refresh=1")
+	if refreshed.Code != http.StatusOK {
+		t.Fatalf("refresh /api/models = %d %s", refreshed.Code, refreshed.Body.String())
+	}
+	if catalog.hits.Load() != 3 {
+		t.Fatalf("refresh=1 hits = %d, want 3", catalog.hits.Load())
+	}
+}
+
 func TestLegacyDiagnosticRoutesAreRemoved(t *testing.T) {
 	srv := New(config.Config{
 		Host:        "127.0.0.1",

internal/api/chat.go

@@ -543,9 +543,16 @@ func (s *Server) filterModelsForIdentity(r *http.Request, models []map[string]an
 	return filtered
 }
 
+const modelsAPICacheTTL = 5 * time.Minute
+
+type modelsAPICacheEntry struct {
+	models []map[string]any
+	at     time.Time
+}
+
 func (s *Server) handleModelsAPI(w http.ResponseWriter, r *http.Request) {
 	refresh := r.URL.Query().Get("refresh") == "1"
-	models, err := s.fetchWorkerModelsFor(refresh, s.requestedAccount(r))
+	models, err := s.fetchModelsAPI(refresh, s.requestedAccount(r))
 	if err != nil {
 		// 503, not 502: some reverse proxies replace origin 502 JSON with
 		// their own HTML error page, which the console then renders as the
@@ -559,6 +566,53 @@ func (s *Server) handleModelsAPI(w http.ResponseWriter, r *http.Request) {
 	})
 }
 
+func modelsAPICacheKey(accountID string) string {
+	if strings.TrimSpace(accountID) == "" {
+		return "*"
+	}
+	return accountID
+}
+
+func cloneModelList(models []map[string]any) []map[string]any {
+	if models == nil {
+		return nil
+	}
+	out := make([]map[string]any, 0, len(models))
+	for _, model := range models {
+		item := make(map[string]any, len(model))
+		for key, value := range model {
+			item[key] = value
+		}
+		out = append(out, item)
+	}
+	return out
+}
+
+// fetchModelsAPI serves GET /api/models from a 5-minute snapshot. Overview and
+// /v1/models keep calling fetchWorkerModelsFor directly so they stay live.
+func (s *Server) fetchModelsAPI(refresh bool, accountID string) ([]map[string]any, error) {
+	key := modelsAPICacheKey(accountID)
+	if !refresh {
+		s.modelsAPICacheMu.Lock()
+		entry, ok := s.modelsAPICache[key]
+		s.modelsAPICacheMu.Unlock()
+		if ok && time.Since(entry.at) < modelsAPICacheTTL {
+			return cloneModelList(entry.models), nil
+		}
+	}
+	models, err := s.fetchWorkerModelsFor(refresh, accountID)
+	if err != nil {
+		return nil, err
+	}
+	s.modelsAPICacheMu.Lock()
+	if s.modelsAPICache == nil {
+		s.modelsAPICache = map[string]modelsAPICacheEntry{}
+	}
+	s.modelsAPICache[key] = modelsAPICacheEntry{models: cloneModelList(models), at: time.Now()}
+	s.modelsAPICacheMu.Unlock()
+	return cloneModelList(models), nil
+}
+
 func (s *Server) handleChatCompletions(w http.ResponseWriter, r *http.Request) {
 	if r.Method != http.MethodPost {
 		writeErr(w, http.StatusMethodNotAllowed, "method_not_allowed", "POST only")

internal/api/server.go

@@ -50,6 +50,8 @@ type Server struct {
 	updateRunning          atomic.Bool
 	updateMu               sync.Mutex
 	updateJob              *systemUpdateJob
+	modelsAPICacheMu       sync.Mutex
+	modelsAPICache         map[string]modelsAPICacheEntry
 }
 
 func New(cfg config.Config) *Server {