Vision model support for Qwen-OAuth (#525)

* refactor: openaiContentGenerator

* refactor: optimize stream handling

* refactor: re-organize refactored files

* fix: unit test cases

* feat: `/model` command for switching to vision model

* fix: lint error

* feat: add image tokenizer to fit vlm context window

* fix: lint and type errors

* feat: add `visionModelPreview` to control default visibility of vision models

* fix: remove deprecated files

* fix: align supported image formats with bailian doc

packages/cli/src/services/BuiltinCommandLoader.test.ts

@@ -56,6 +56,13 @@ vi.mock('../ui/commands/mcpCommand.js', () => ({
     kind: 'BUILT_IN',
   },
 }));
+vi.mock('../ui/commands/modelCommand.js', () => ({
+  modelCommand: {
+    name: 'model',
+    description: 'Model command',
+    kind: 'BUILT_IN',
+  },
+}));
 
 describe('BuiltinCommandLoader', () => {
   let mockConfig: Config;
@@ -126,5 +133,8 @@ describe('BuiltinCommandLoader', () => {
 
     const mcpCmd = commands.find((c) => c.name === 'mcp');
     expect(mcpCmd).toBeDefined();
+
+    const modelCmd = commands.find((c) => c.name === 'model');
+    expect(modelCmd).toBeDefined();
   });
 });