feat: add interactive setup wizard and model error handling

Add comprehensive setup command with provider selection, API key
configuration, and model selection. Include error recovery for
model-not-found scenarios with suggested alternatives and automatic
retry functionality. Update Anthropic model list with latest versions
and add provider metadata for better user experience.

src/engine/gemini.ts

@@ -7,6 +7,7 @@ import {
 } from '@google/generative-ai';
 import axios from 'axios';
 import { OpenAI } from 'openai';
+import { ModelNotFoundError } from '../utils/errors';
 import { removeContentTags } from '../utils/removeContentTags';
 import { AiEngine, AiEngineConfig } from './Engine';
 
@@ -76,6 +77,15 @@ export class GeminiEngine implements AiEngine {
       return removeContentTags(content, 'think');
     } catch (error) {
       const err = error as Error;
+
+      // Check for model not found errors
+      if (err.message?.toLowerCase().includes('model') &&
+          (err.message?.toLowerCase().includes('not found') ||
+           err.message?.toLowerCase().includes('does not exist') ||
+           err.message?.toLowerCase().includes('invalid'))) {
+        throw new ModelNotFoundError(this.config.model, 'gemini', 404);
+      }
+
       if (
         axios.isAxiosError<{ error?: { message: string } }>(error) &&
         error.response?.status === 401
@@ -84,6 +94,11 @@ export class GeminiEngine implements AiEngine {
         if (geminiError) throw new Error(geminiError?.message);
       }
 
+      // Check axios 404 errors
+      if (axios.isAxiosError(error) && error.response?.status === 404) {
+        throw new ModelNotFoundError(this.config.model, 'gemini', 404);
+      }
+
       throw err;
     }
   }