Lovable Prompt Creator - AI Prompt Generator

```markdown
# PROJECT DEFINITION
Build a comprehensive web application named "Figaro" that automates the recreation of sales call recordings using AI-generated video and voice. The application will allow sales representatives to receive coaching videos based on their calls, showcasing ideal dialogues, and providing feedback through a structured pipeline. The objective is to improve sales performance by delivering personalized coaching material in a user-friendly interface.

# DESIGN REQUIREMENTS (include these exact specs—typography, primary color, border radius—in your output):
- Visual Style: minimalist - Clean, simple design with plenty of white space, minimal color palette, and focus on typography
- Color Mode: Light theme with dark text on light backgrounds
- Primary Color: #1978E5 (accent for buttons, links, highlights)
- Typography: Use Inter from Google Fonts for headings, Inter for body text and UI elements
- Border Radius: 8px (moderately rounded) for buttons, cards, and inputs

# TECH STACK & CONSTRAINTS
- Language: JavaScript (Node.js with Express framework)
- Database: Supabase (Postgres with storage and real-time capabilities)
- Job scheduling: node-cron for polling jobs
- Video processing: FFmpeg for video compositing
- APIs: Integrate Claude AI, HeyGen, ElevenLabs, HubSpot, and Slack using RESTful calls via fetch
- Ensure minimal dependencies by avoiding specific SDK installations and using plain fetch requests instead.

# BREAKDOWN INTO TASKS
1. **Project Initialization**
   - Set up a new Replit project with Node.js and install required packages.
   - Configure `replit.nix` for system dependencies (Node.js, FFmpeg).
  
2. **Database Setup**
   - Create the necessary tables in Supabase using the provided SQL schema.
   - Seed initial data for representatives.

3. **API Integration**
   - Implement API calls for Claude AI, HeyGen, ElevenLabs, HubSpot, and Slack in their respective service files.
   - Create wrappers for API interaction to streamline calls.

4. **Pipeline Logic**
   - Develop the main job processor in `workers/pipeline.js` to handle job status transitions.
   - Implement all steps of the pipeline (extracting, scoring, rewriting, generating avatar, composing, delivering) in the designated folders.

5. **Webhooks and API Routes**
   - Create routes for handling webhooks from HubSpot and API endpoints for managing jobs and reps.
   - Ensure the integration with Slack for video delivery and notifications.

6. **Frontend Development**
   - Build a simple dashboard using HTML, CSS, and JavaScript for managing jobs, representatives, and manual triggers.
   - Implement real-time updates using Supabase's client library.

7. **Testing and Validation**
   - Create test jobs and validate each step of the pipeline.
   - Ensure that video processing and API interactions function as expected.

# IMPLEMENTATION GUIDANCE
- Start by setting up the server and defining the necessary routes.
- Implement the job processing logic step by step, testing each component as it is developed.
- Use console logging extensively to debug and verify API responses and internal states.
- Create a user-friendly dashboard to visualize job statuses and allow manual interventions.

# ITERATION STRATEGY
- After completing the initial implementation, gather feedback from users to identify pain points.
- Iterate the design based on user interactions, particularly focusing on the dashboard's usability and the clarity of feedback provided.

# TESTING & VALIDATION
- Perform end-to-end tests by simulating real sales calls and verifying that the entire pipeline from recording to Slack delivery works seamlessly.
- Validate the correctness of AI-generated content by comparing it against expected outputs.
- Monitor logs for errors and optimize the performance of the video processing steps.
```