Can I vibecode Tweet Hunter?

Build a personal X (Twitter) growth planner inspired by Tweet Hunter.
Use exactly this stack: Next.js 15 (App Router), TypeScript, Tailwind CSS, PostgreSQL (via Drizzle ORM), and BullMQ (for background scheduling).

DATA MODEL:
- `Draft`: id, content, media_urls (json), is_thread (boolean), thread_parent_id, created_at, updated_at
- `ScheduledPost`: id, draft_id, scheduled_for, status (pending, published, failed), platform_post_id, error_message
- `PerformanceData`: id, post_id, impressions, likes, retweets, replies, fetched_at

USER JOURNEYS & SCREENS:
1. **Composer Screen (`/compose`)**: A distraction-free writing interface mimicking X's UI. Include a character counter (280 limit) and a "Add to Thread" button. Show a live preview card simulating the mobile X feed. 
2. **Inspiration/Swipe File (`/inspiration`)**: A grid displaying high-performing example tweets (pre-seeded in the DB) to overcome writer's block. Include a "Use as Template" button that copies the structure to the Composer.
3. **Queue/Calendar (`/queue`)**: A weekly calendar view showing all `ScheduledPost` records. Allow drag-and-drop to change the `scheduled_for` time. 
4. **Analytics Dashboard (`/analytics`)**: A simple chart (use Recharts) displaying `PerformanceData` over time, highlighting top-performing tweets based on engagement rate.

TECHNICAL SPECIFICATIONS:
- Scheduling Engine: Use Redis and BullMQ to handle the scheduling queue. A worker process should poll for `ScheduledPost` entries where `scheduled_for` <= now and `status` == 'pending'.
- X API Integration: Use the official X API v2 (via `twitter-api-v2` npm package) to publish tweets. Implement OAuth 2.0 PKCE flow for user authentication and token storage.
- Auto-Plug: Implement a feature in the Composer to automatically add a reply (e.g., a newsletter link) to a thread if the primary tweet hits a certain engagement threshold (requires periodic polling of X API).
- AI Generation: Integrate OpenAI API (gpt-4o-mini) to provide a "Rewrite" button in the Composer (e.g., make it punchier, professional, or engaging).

Make the default mode single-user and private.
Store all user data locally.
Put X API credentials and OpenAI keys in .env and provide .env.example.
Include clear empty, loading, validation, success, and failure states.
Use accessible keyboard navigation.
Write focused unit tests for the data model and the scheduling workflow.
Create a README with setup, permissions, architecture, data location, and run scripts.
Do not add third-party analytics or ads.