Can I vibecode VerifiedDR?

Build me a sophisticated, local search and AI visibility tracker inspired by VerifiedDR, explicitly tailored for a single domain. 

## Tech Stack & Architecture
- **Backend & Database:** Node.js (v22), Express, and better-sqlite3. The app should bind exclusively to localhost:4173.
- **Frontend & UI:** Server-rendered HTML powered by a templating engine (e.g., EJS or Pug) mixed with vanilla JavaScript. Use Tailwind CSS via CDN or raw CSS for a premium, sleek aesthetic—incorporating a dark mode by default, glassmorphism elements, subtle hover micro-animations, and dynamic sparklines.
- **Configuration:** Maintain all API secrets (OpenAI, Perplexity, Gemini, Google OAuth) in a secure `.env` file. Do not implement multi-user accounts or telemetry to ensure complete privacy.

## Data Models & State
Create a robust SQLite schema:
1.  **Settings Table:** Stores the target domain name, a list of brand aliases, and a curated list of up to 20 buyer-intent questions/queries.
2.  **GSC Metrics Table:** Stores daily metrics (clicks, impressions, CTR, average position) imported from Google Search Console, keyed by date, query, and landing page.
3.  **AI Mentions Table:** Stores historical responses to the 20 questions across multiple AI models. Track run date, full model answer, citations/URLs, and boolean flags for brand/domain mentions.
4.  **Crawl Data Table:** Stores weekly sitemap crawl results (URL, status code, meta title, canonical URL, noindex flag, and broken internal links).

## Core Features & Logic
1.  **Google Search Console Integration:** Use Google OAuth to authenticate. Implement a sync job that fetches the last 28 days of performance data (clicks, impressions, CTR, average position).
2.  **AI Visibility Tracking:** On-demand or scheduled weekly, query the OpenAI, Perplexity, and Gemini APIs with the 20 buyer questions. Parse the responses to determine if the target domain or brand aliases are mentioned and cited.
3.  **Site Health Crawler:** A weekly cron job (using node-cron, undici, and cheerio) that parses the `robots.txt` and `sitemap.xml`, crawling all pages to check for 404s, missing titles, canonical mismatches, and broken internal links.
4.  **Local Dashboard & Action Queue:**
    -   **Overview:** Display a high-level summary showing 7-day and 28-day GSC changes, AI mention rate percentages per model, and top cited competitor domains.
    -   **Sparklines:** Render historical trend sparklines for each of the 20 questions to visualize visibility over time.
    -   **Action Queue:** A prioritized inbox of tasks for the webmaster: (1) Lost AI mentions (a model used to cite the site but stopped), (2) Pages with falling GSC traffic/rankings, (3) Technical SEO errors (broken links, unexpected noindex tags). Each task must link directly to the underlying evidence.
5.  **Export & Retention:** Store raw JSON responses for debugging and export every scheduled run as both a detailed Markdown report and a CSV file saved locally in a `./exports/` directory.

## Documentation
Include a comprehensive `README.md` detailing the Google OAuth setup process, required API keys, estimated monthly API costs based on token usage, SQLite backup/restore commands, and instructions for triggering a manual sync.