fix: normalize catalog provider ids for probe model selection

src/commands/models/list.probe.targets.test.ts

@@ -1,10 +1,12 @@
 import { beforeEach, describe, expect, it, vi } from "vitest";
 import type { AuthProfileStore } from "../../agents/auth-profiles.js";
 import { OLLAMA_LOCAL_AUTH_MARKER } from "../../agents/model-auth-markers.js";
+import type { ModelCatalogEntry } from "../../agents/model-catalog.js";
 import type { OpenClawConfig } from "../../config/config.js";
 
 let mockStore: AuthProfileStore;
 let mockAllowedProfiles: string[];
+const loadModelCatalogMock = vi.fn<() => Promise<ModelCatalogEntry[]>>(async () => []);
 
 const resolveAuthProfileOrderMock = vi.fn(() => mockAllowedProfiles);
 const resolveAuthProfileEligibilityMock = vi.fn(() => ({
@@ -14,7 +16,7 @@ const resolveAuthProfileEligibilityMock = vi.fn(() => ({
 const resolveSecretRefStringMock = vi.fn(async () => "resolved-secret");
 
 vi.mock("../../agents/model-catalog.js", () => ({
-  loadModelCatalog: vi.fn(async () => []),
+  loadModelCatalog: loadModelCatalogMock,
 }));
 vi.mock("../../secrets/resolve.js", () => ({
   resolveSecretRefString: resolveSecretRefStringMock,
@@ -130,6 +132,8 @@ describe("buildProbeTargets reason codes", () => {
       },
     };
     mockAllowedProfiles = [];
+    loadModelCatalogMock.mockReset();
+    loadModelCatalogMock.mockResolvedValue([]);
     resolveAuthProfileOrderMock.mockClear();
     resolveAuthProfileEligibilityMock.mockClear();
     resolveSecretRefStringMock.mockReset();
@@ -217,4 +221,48 @@ describe("buildProbeTargets reason codes", () => {
       );
     });
   });
+
+  it("matches canonical providers against alias-valued catalog probe models", async () => {
+    mockStore = {
+      version: 1,
+      profiles: {},
+      order: {},
+    };
+    loadModelCatalogMock.mockResolvedValueOnce([
+      { provider: "z.ai", id: "glm-4.7", name: "GLM-4.7" },
+    ]);
+
+    const plan = await buildProbeTargets({
+      cfg: {
+        models: {
+          providers: {
+            zai: {
+              baseUrl: "https://api.z.ai/v1",
+              api: "openai-responses",
+              apiKey: "sk-zai-test", // pragma: allowlist secret
+              models: [],
+            },
+          },
+        },
+      } as OpenClawConfig,
+      providers: ["zai"],
+      modelCandidates: [],
+      options: {
+        timeoutMs: 5_000,
+        concurrency: 1,
+        maxTokens: 16,
+      },
+    });
+
+    expect(plan.results).toEqual([]);
+    expect(plan.targets).toHaveLength(1);
+    expect(plan.targets[0]).toEqual(
+      expect.objectContaining({
+        provider: "zai",
+        model: { provider: "zai", model: "glm-4.7" },
+        source: "models.json",
+        label: "models.json",
+      }),
+    );
+  });
 });

src/commands/models/list.probe.ts

@@ -147,9 +147,9 @@ function selectProbeModel(params: {
   if (direct && direct.length > 0) {
     return { provider, model: direct[0] };
   }
-  const fromCatalog = catalog.find((entry) => entry.provider === provider);
+  const fromCatalog = catalog.find((entry) => normalizeProviderId(entry.provider) === provider);
   if (fromCatalog) {
-    return { provider: fromCatalog.provider, model: fromCatalog.id };
+    return { provider, model: fromCatalog.id };
   }
   return null;
 }