script for internal personal use. VPS is already available

Technical assignment: monitoring Truth Social

1. Project description

Automated account monitoring system on the Truth Social platform with phone notifications and AI content analysis.

2. Functional requirements

2.1 Monitoring

• Check for new posts from the specified account
• Check frequency: 4 times per second
• Use of proxy servers to bypass restrictions
• Automatic IP address rotation

2.2 Notifications

• Phone call when a new post appears
• Voice message with customizable text
• Uses Twilio API

2.3 Content analysis

• Extract text from the new post
• Send text to DeepSeek AI for analysis
• Customizable prompt for analysis
• Save results to the database

2.4 Web interface

• View the list of recent posts
• Display AI analysis results
• System status (active/stopped)
• Time of last check

3. Technologies

Backend:

• Python 3.10+
• Flask (web server)
• cloudscraper (bypass Cloudflare protection)
• twilio (phone calls)
• openai (compatible with DeepSeek API)
• sqlite3 (database)

Frontend:

• HTML
• Bootstrap 5 (via CDN)
• JavaScript (data updates)

4. Project structure

project/

│

├── config.json          # Configuration file

├── app.py              # Main application code

├── templates/

│   └── index.html      # Web interface

└── data.db             # Database (created automatically)

5. Configuration file (config.json)

{

  "target_account": "username",

  "scan_frequency": 4,

  

  "proxies": [

  

  ],

  

  "twilio": {

    "account_sid": "ACxxxxxxxxxxxxx",

    "auth_token": "your_auth_token",

    "from_number": "+",

    "to_number": "+",

    "message": "New post detected"

  },

  

  "deepseek": {

    "api_key": "sk-xxxxxxxxxxxxx",

    "prompt": "Analyze the following post and determine: sentiment (positive/negative/neutral), main topics, mentions of companies or countries. Post: {text}"

  }

}

6. Database

Table posts:

- id (PRIMARY KEY)

- post_id (unique post ID)

- username (account name)

- text (post text)

- created_at (publication time)

- ai_analysis (analysis result from DeepSeek)

- call_status (call status: success/error)

7. Architecture

Web interface (Flask)

    ↓

    ├── GET /          → Main page with post table

    ├── GET /api/posts → JSON with the list of posts

    └── GET /api/status → JSON system status



Background process (separate thread):

    

    [Check account] 

           ↓

    [New post found?]

           ↓ (yes)

    [Call via Twilio]

           ↓

    [Analyze via DeepSeek]

           ↓

    [Save to SQLite]

           ↓

    [Wait 0.25 sec] → [Repeat]

8. Workflow algorithm

1. Start application

   • Read config.json
   • Connect to the database
   • Start Flask server
   • Start background monitoring thread

2. Monitoring loop (every 0.25 seconds):

   • Select the next proxy from the list
   • Request the account page
   • Extract the ID of the last post
   • Compare with the previous ID
   • If new post → start processing

3. Process new post:

   • Extract post text
   • Call via Twilio
   • Send text to DeepSeek
   • Receive analysis result
   • Save to database

4. Web interface:

   • JavaScript requests /api/posts every 0.25 seconds
   • Update post table
   • Display system status

9. Web interface

Main page contains:

• Header with system name
• Status indicator (green = running)
• Time of last check
• Table with columns:
  • Publication time
  • Account name
  • Post text (first 200 characters)
  • AI analysis result
  • Call status

10. Installation and launch

Step 1. Install dependencies:

pip install flask cloudscraper twilio openai

Step 2. Configure config.json:

• Specify the account name for monitoring
• Add a list of proxy servers
• Enter Twilio account details
• Add DeepSeek API key
• Configure prompt for analysis

Step 3. Launch:

python app.py

Step 4. Open browser:

http://localhost:5000

11. Server requirements

Minimum:

• Python 3.10+
• 512 MB RAM
• 5 GB disk space
• Stable internet connection

Recommended:

• VPS/Cloud server (Digital Ocean, AWS, etc.)
• 1 GB RAM
• Ubuntu 22.04 or similar OS

12. Complexity assessment

Development complexity: 3/10

Development time:

• From scratch: 4-5 hours
• With AI assistance: 2-3 hours

Components:

• ✅ All libraries are well documented
• ✅ Simple architecture (one file)
• ✅ Standard APIs (Twilio, DeepSeek)
• ⚠️ Only complexity: bypassing Cloudflare

13. What AI can write

The neural network can handle the task 95%:

• ✅ Flask application structure
• ✅ Twilio API integration
• ✅ DeepSeek API integration
• ✅ Working with SQLite
• ✅ HTML interface
• ✅ Monitoring logic
• ⚠️ Manual configuration for bypassing Cloudflare may be required

14. Possible problems and solutions

Problem: Cloudflare blocks requests
Solution: Use quality proxies, increase pauses between requests

Problem: Proxies do not work
Solution: Regularly update the list of proxies in config.json

Problem: DeepSeek returns an error
Solution: Check API key and account balance

Problem: Twilio does not call
Solution: Check account balance, number verification

Conclusion: The project is technically feasible within a few hours using standard tools. The main code can be generated by an AI assistant with minimal adjustments.