fix: cron model fallback to agent defaults when payload.model fails (#26717)

Merged via /review-pr -> /prepare-pr -> /merge-pr.

Prepared head SHA: 06454bd55b
Co-authored-by: Youyou972 <50808411+Youyou972@users.noreply.github.com>
Co-authored-by: shakkernerd <165377636+shakkernerd@users.noreply.github.com>
Reviewed-by: @shakkernerd

CHANGELOG.md

@@ -16,6 +16,7 @@ Docs: https://docs.openclaw.ai
 - Security/Nextcloud Talk: drop replayed signed webhook events with persistent per-account replay dedupe across restarts, and reject unexpected webhook backend origins when account base URL is configured. Thanks @aristorechina for reporting.
 - Security/Gateway: harden `agents.files` path handling to block out-of-workspace symlink targets for `agents.files.get`/`agents.files.set`, keep in-workspace symlink targets supported, and add gateway regression coverage for both blocked escapes and allowed in-workspace symlinks. Thanks @tdjackey for reporting.
 - Gateway/Message media roots: thread `agentId` through gateway `send` RPC and prefer explicit `agentId` over session/default resolution so non-default agent workspace media sends no longer fail with `LocalMediaAccessError`; added regression coverage for agent precedence and blank-agent fallback. (#23249) Thanks @Sid-Qin.
+- Cron/Model overrides: when isolated `payload.model` is no longer allowlisted, fall back to default model selection instead of failing the job, while still returning explicit errors for invalid model strings. (#26717) Thanks @Youyou972.
 - Security/Nextcloud Talk: reject unsigned webhook traffic before full body reads, reducing unauthenticated request-body exposure, with auth-order regression coverage. (#26118) Thanks @bmendonca3.
 - Security/Nextcloud Talk: stop treating DM pairing-store entries as group allowlist senders, so group authorization remains bounded to configured group allowlists. (#26116) Thanks @bmendonca3.
 - Security/IRC: keep pairing-store approvals DM-only and out of IRC group allowlist authorization, with policy regression tests for allowlist resolution. (#26112) Thanks @bmendonca3.

src/cron/isolated-agent/run.skill-filter.test.ts

@@ -6,6 +6,13 @@ import { runWithModelFallback } from "../../agents/model-fallback.js";
 const buildWorkspaceSkillSnapshotMock = vi.fn();
 const resolveAgentConfigMock = vi.fn();
 const resolveAgentSkillsFilterMock = vi.fn();
+const getModelRefStatusMock = vi.fn().mockReturnValue({ allowed: false });
+const isCliProviderMock = vi.fn().mockReturnValue(false);
+const resolveAllowedModelRefMock = vi.fn();
+const resolveConfiguredModelRefMock = vi.fn();
+const resolveHooksGmailModelMock = vi.fn();
+const resolveThinkingDefaultMock = vi.fn();
+const logWarnMock = vi.fn();
 
 vi.mock("../../agents/agent-scope.js", () => ({
   resolveAgentConfig: resolveAgentConfigMock,
@@ -36,14 +43,12 @@ vi.mock("../../agents/model-selection.js", async (importOriginal) => {
   const actual = await importOriginal<typeof import("../../agents/model-selection.js")>();
   return {
     ...actual,
-    getModelRefStatus: vi.fn().mockReturnValue({ allowed: false }),
+    getModelRefStatus: getModelRefStatusMock,
-    isCliProvider: vi.fn().mockReturnValue(false),
+    isCliProvider: isCliProviderMock,
-    resolveAllowedModelRef: vi
+    resolveAllowedModelRef: resolveAllowedModelRefMock,
-      .fn()
+    resolveConfiguredModelRef: resolveConfiguredModelRefMock,
-      .mockReturnValue({ ref: { provider: "openai", model: "gpt-4" } }),
+    resolveHooksGmailModel: resolveHooksGmailModelMock,
-    resolveConfiguredModelRef: vi.fn().mockReturnValue({ provider: "openai", model: "gpt-4" }),
+    resolveThinkingDefault: resolveThinkingDefaultMock,
-    resolveHooksGmailModel: vi.fn().mockReturnValue(null),
-    resolveThinkingDefault: vi.fn().mockReturnValue(undefined),
   };
 });
 
@@ -138,7 +143,7 @@ vi.mock("../../infra/skills-remote.js", () => ({
 }));
 
 vi.mock("../../logger.js", () => ({
-  logWarn: vi.fn(),
+  logWarn: (...args: unknown[]) => logWarnMock(...args),
 }));
 
 vi.mock("../../security/external-content.js", () => ({
@@ -222,6 +227,13 @@ describe("runCronIsolatedAgentTurn — skill filter", () => {
     });
     resolveAgentConfigMock.mockReturnValue(undefined);
     resolveAgentSkillsFilterMock.mockReturnValue(undefined);
+    resolveConfiguredModelRefMock.mockReturnValue({ provider: "openai", model: "gpt-4" });
+    resolveAllowedModelRefMock.mockReturnValue({ ref: { provider: "openai", model: "gpt-4" } });
+    resolveHooksGmailModelMock.mockReturnValue(null);
+    resolveThinkingDefaultMock.mockReturnValue(undefined);
+    getModelRefStatusMock.mockReturnValue({ allowed: false });
+    isCliProviderMock.mockReturnValue(false);
+    logWarnMock.mockReset();
     // Fresh session object per test — prevents mutation leaking between tests
     resolveCronSessionMock.mockReturnValue({
       storePath: "/tmp/store.json",
@@ -408,5 +420,78 @@ describe("runCronIsolatedAgentTurn — skill filter", () => {
     it("preserves defaults when agent overrides primary in object form", async () => {
       await expectPrimaryOverridePreservesDefaults({ primary: "anthropic/claude-sonnet-4-5" });
     });
+
+    it("applies payload.model override when model is allowed", async () => {
+      resolveAllowedModelRefMock.mockReturnValueOnce({
+        ref: { provider: "anthropic", model: "claude-sonnet-4-6" },
+      });
+
+      const result = await runCronIsolatedAgentTurn(
+        makeParams({
+          job: makeJob({
+            payload: { kind: "agentTurn", message: "test", model: "anthropic/claude-sonnet-4-6" },
+          }),
+        }),
+      );
+
+      expect(result.status).toBe("ok");
+      expect(logWarnMock).not.toHaveBeenCalled();
+      expect(runWithModelFallbackMock).toHaveBeenCalledOnce();
+      const runParams = runWithModelFallbackMock.mock.calls[0][0];
+      expect(runParams.provider).toBe("anthropic");
+      expect(runParams.model).toBe("claude-sonnet-4-6");
+    });
+
+    it("falls back to agent defaults when payload.model is not allowed", async () => {
+      resolveAllowedModelRefMock.mockReturnValueOnce({
+        error: "model not allowed: anthropic/claude-sonnet-4-6",
+      });
+
+      const result = await runCronIsolatedAgentTurn(
+        makeParams({
+          cfg: {
+            agents: {
+              defaults: {
+                model: { primary: "openai-codex/gpt-5.3-codex", fallbacks: defaultFallbacks },
+              },
+            },
+          },
+          job: makeJob({
+            payload: { kind: "agentTurn", message: "test", model: "anthropic/claude-sonnet-4-6" },
+          }),
+        }),
+      );
+
+      expect(result.status).toBe("ok");
+      expect(logWarnMock).toHaveBeenCalledWith(
+        "cron: payload.model 'anthropic/claude-sonnet-4-6' not allowed, falling back to agent defaults",
+      );
+      expect(runWithModelFallbackMock).toHaveBeenCalledOnce();
+      const callCfg = runWithModelFallbackMock.mock.calls[0][0].cfg;
+      const model = callCfg?.agents?.defaults?.model as
+        | { primary?: string; fallbacks?: string[] }
+        | undefined;
+      expect(model?.primary).toBe("openai-codex/gpt-5.3-codex");
+      expect(model?.fallbacks).toEqual(defaultFallbacks);
+    });
+
+    it("returns an error when payload.model is invalid", async () => {
+      resolveAllowedModelRefMock.mockReturnValueOnce({
+        error: "invalid model: openai/",
+      });
+
+      const result = await runCronIsolatedAgentTurn(
+        makeParams({
+          job: makeJob({
+            payload: { kind: "agentTurn", message: "test", model: "openai/" },
+          }),
+        }),
+      );
+
+      expect(result.status).toBe("error");
+      expect(result.error).toBe("invalid model: openai/");
+      expect(logWarnMock).not.toHaveBeenCalled();
+      expect(runWithModelFallbackMock).not.toHaveBeenCalled();
+    });
   });
 });

src/cron/isolated-agent/run.ts

@@ -198,10 +198,17 @@ export async function runCronIsolatedAgentTurn(params: {
       defaultModel: resolvedDefault.model,
     });
     if ("error" in resolvedOverride) {
-      return { status: "error", error: resolvedOverride.error };
+      if (resolvedOverride.error.startsWith("model not allowed:")) {
+        logWarn(
+          `cron: payload.model '${modelOverride}' not allowed, falling back to agent defaults`,
+        );
+      } else {
+        return { status: "error", error: resolvedOverride.error };
+      }
+    } else {
+      provider = resolvedOverride.ref.provider;
+      model = resolvedOverride.ref.model;
     }
-    provider = resolvedOverride.ref.provider;
-    model = resolvedOverride.ref.model;
   }
   const now = Date.now();
   const cronSession = resolveCronSession({