Can I vibecode Mixo?

Build a usable personal replacement for the core loop of Mixo, a platform for launching websites in seconds with AI.
Use exactly this stack: Next.js 15, React, TypeScript, Tailwind CSS, and SQLite for local storage.
Primary job: Build an AI-powered website builder that allows users to instantly generate a full website or landing page by simply providing a short text description of their business idea or concept.

Core functionalities to implement:
1. AI-Powered Generation: 
   - Accept a short text prompt (e.g., "A modern coffee shop in Brooklyn") and integrate with the OpenAI or Anthropic API to generate structured web content, compelling copy, and suitable page layouts tailored to the niche.
   - Automatically generate the hero section, features list, testimonials, and a call-to-action (CTA).
2. No-Code Editor & Customization:
   - Provide an intuitive, lightweight WYSIWYG or drag-and-drop editor interface.
   - Allow users to customize aesthetics: edit text, swap placeholder images, change brand colors, and update fonts.
   - Ensure all generated designs are inherently mobile-responsive and look great on all devices out of the box.
3. Marketing & Lead Generation Tools:
   - Include built-in lead generation features: a functional email capture form and waitlist management system that stores submissions in the local SQLite database.
   - Implement basic SEO features (auto-generating title tags and meta descriptions based on the site's content).
4. Publishing & Preview:
   - Provide a real-time preview of the responsive layout.
   - Implement a mechanism to "publish" or export the site as static HTML/CSS files for easy hosting anywhere.

General requirements:
- Start from an empty folder and create the complete working project.
- Make the default mode single-user and private. Store user data and generated sites locally using SQLite.
- Put every secret and external credential (like the LLM API key) in .env and provide .env.example.
- Include realistic sample data and placeholder images for sites when first loaded.
- Implement the smallest polished interface that completes the core loop end-to-end with clear empty, loading, validation, success, and failure states.
- Deliberately exclude deep customization capabilities (like a full Webflow-style visual editor) or massive component libraries. Focus on speed and simplicity.
- Ensure the app is resilient: if the LLM API is unavailable, the user can still manually create a site using a default blank template and the editor.
- Write focused unit tests for the database model (sites, leads) and the core generation workflow.
- Add an end-to-end smoke test that proves a site can be created, edited, and lead captured.
- Create a detailed README with setup instructions, architecture, and limitations.
- Run tests and build before finishing, fixing any errors rather than merely describing them.