feat: add City10 platform with Laravel 13, AI agents, 3D viewer, and LGPD compliance

Initial commit of the complete City10 virtual city simulation platform.

Backend (Laravel 13):
- REST API with Sanctum tokens and session-based web authentication
- Role-based access control (admin/user) with middleware protection
- AI agent CRUD with OpenRouter.ai integration restricted to free models
- Store management with 3D coordinates and AI shopkeeper assignment
- Versioned legal documents (Terms, Privacy, Cookies) with LGPD audit trail
- Math CAPTCHA on public forms and multi-language support (pt_BR, en, es)

Frontend:
- Responsive Blade templates with Tailwind CSS v4
- Three.js 3D city viewer with SSE chat streaming to AI agents
- Admin panel for store and user administration

Infrastructure:
- MariaDB/MySQL database with comprehensive migrations
- Database seeders for demo content and test accounts
- PHPUnit test suite and Laravel Pint code styling
- Vite build configuration with hot reload

config/ai.php

@@ -0,0 +1,78 @@
+<?php
+
+return [
+
+    /*
+    |--------------------------------------------------------------------------
+    | OpenRouter.ai Integration
+    |--------------------------------------------------------------------------
+    |
+    | Credentials and behaviour settings used by the AI agents to take
+    | strategic decisions through the OpenRouter.ai chat completions API.
+    | The API key must be provided via the OPENROUTER_API_KEY env variable
+    | and must never be committed to version control.
+    |
+    */
+
+    'openrouter' => [
+        'api_key' => env('OPENROUTER_API_KEY'),
+        'base_url' => env('OPENROUTER_BASE_URL', 'https://openrouter.ai/api/v1'),
+        'timeout' => (int) env('OPENROUTER_TIMEOUT', 30),
+        'app_name' => env('APP_NAME', 'City10'),
+        'app_url' => env('APP_URL', 'http://localhost:16980'),
+    ],
+
+    /*
+    |--------------------------------------------------------------------------
+    | Allow-listed Free Models
+    |--------------------------------------------------------------------------
+    |
+    | Only the free models listed below may be selected for AI agents.
+    | Administrators cannot register agents with models outside this list,
+    | which keeps operating costs at zero and enforces the platform policy.
+    |
+    */
+
+    'allowed_models' => [
+        'poolside/laguna-m.1:free',
+        'cohere/north-mini-code:free',
+        'deepseek/deepseek-v4-flash-0731:free',
+        'qwen/qwen3.8-27b:free',
+        'nvidia/nemotron-3-ultra-550b-a55b:free',
+        'nvidia/nemotron-3-super-120b-a12b:free',
+        'google/gemma-4-31b-it:free',
+        'google/gemma-4-26b-a4b-it:free',
+        'openai/gpt-oss-20b:free',
+        'nvidia/nemotron-nano-12b-v2-vl:free',
+        'nvidia/nemotron-3-nano-omni-30b-a3b-reasoning:free',
+        'nvidia/nemotron-3-nano-30b-a3b:free',
+        'nvidia/nemotron-nano-9b-v2:free',
+        'poolside/laguna-xs-2.1:free',
+        'inclusionai/ling-3.0-flash-vl:free',
+        'qwen/qwen3.7-flash:free',
+    ],
+
+    /*
+    |--------------------------------------------------------------------------
+    | Default Model
+    |--------------------------------------------------------------------------
+    |
+    | Model assigned to newly created agents when none is specified.
+    |
+    */
+
+    'default_model' => env('AI_DEFAULT_MODEL', 'deepseek/deepseek-v4-flash-0731'),
+
+    /*
+    |--------------------------------------------------------------------------
+    | Agent Decision Cadence (seconds)
+    |--------------------------------------------------------------------------
+    |
+    | Minimum interval between two reasoning calls for the same agent,
+    | protecting the upstream API from excessive usage.
+    |
+    */
+
+    'think_interval' => (int) env('AI_THINK_INTERVAL', 60),
+
+];