fix: provider-aware model routing with fallback visibility

- Extract createClientFromConfig() to dispatch on provider field instead
  of hardcoding all tiers as AnthropicClient
- Add fallback/fallbackReason metadata to ChatResponse and ChatStreamEvent
  so callers know when a fallback model was used
- Enhance doctor check to report full model stack and warn on missing
  API keys for cloud providers
- Log fallback warnings in NativeAgent and display them in TUI
- Support tier names and local_providers entries in fallback_chain
- Add 8 tests for createClientFromConfig covering all provider types

src/daemon/clientFactory.test.ts

@@ -0,0 +1,78 @@
+import { describe, it, expect } from 'vitest';
+import { createClientFromConfig } from './index.js';
+import { AnthropicClient } from '../models/anthropic.js';
+import { OpenAIClient } from '../models/openai.js';
+import { OllamaClient } from '../models/local/ollama.js';
+import { LlamaCppClient } from '../models/local/llamacpp.js';
+
+describe('createClientFromConfig', () => {
+  it('creates AnthropicClient for anthropic provider', () => {
+    const client = createClientFromConfig({
+      provider: 'anthropic',
+      model: 'claude-sonnet-4-5-20250514',
+      api_key: 'sk-ant-test',
+    });
+    expect(client).toBeInstanceOf(AnthropicClient);
+  });
+
+  it('creates OpenAIClient for openai provider', () => {
+    const client = createClientFromConfig({
+      provider: 'openai',
+      model: 'gpt-4o',
+      api_key: 'sk-test',
+    });
+    expect(client).toBeInstanceOf(OpenAIClient);
+  });
+
+  it('creates OllamaClient for ollama provider', () => {
+    const client = createClientFromConfig({
+      provider: 'ollama',
+      model: 'llama3.2:1b',
+      endpoint: 'http://localhost:11434',
+    });
+    expect(client).toBeInstanceOf(OllamaClient);
+  });
+
+  it('creates OllamaClient with num_gpu option', () => {
+    const client = createClientFromConfig({
+      provider: 'ollama',
+      model: 'llama3.2:1b',
+      num_gpu: 0,
+    });
+    expect(client).toBeInstanceOf(OllamaClient);
+  });
+
+  it('creates LlamaCppClient for llamacpp provider', () => {
+    const client = createClientFromConfig({
+      provider: 'llamacpp',
+      model: 'ministral-reasoning',
+      endpoint: 'http://localhost:8080',
+    });
+    expect(client).toBeInstanceOf(LlamaCppClient);
+  });
+
+  it('defaults llamacpp endpoint to localhost:8080', () => {
+    const client = createClientFromConfig({
+      provider: 'llamacpp',
+      model: 'test-model',
+    });
+    expect(client).toBeInstanceOf(LlamaCppClient);
+  });
+
+  it('creates OpenAI-compatible client for gemini provider (with warning)', () => {
+    const client = createClientFromConfig({
+      provider: 'gemini',
+      model: 'gemini-2.5-pro',
+      api_key: 'test-key',
+    });
+    // Gemini falls back to OpenAI-compatible client
+    expect(client).toBeInstanceOf(OpenAIClient);
+  });
+
+  it('throws for unknown provider', () => {
+    expect(() => createClientFromConfig({
+      provider: 'unknown' as 'anthropic',
+      model: 'test',
+    })).toThrow('Unknown model provider: unknown');
+  });
+});