Fix incorrect Threat Confidence for safe messages

[shrutig-blip]

Summary

This PR fixes the incorrect Threat Confidence displayed for safe messages.

fix issue: #183

Changes

• Added a separate threatConfidence value in the frontend.
• Display 0% Threat Confidence when the classification is safe.
• Continue displaying the reported confidence for suspicious and danger classifications.

Testing

Tested with:

• ✅ "hi"
• ✅ "hello"
• ✅ "How are you?"
• ✅ Fake job template
• ✅ OTP fraud template
• ✅ Phishing message

The issue where harmless messages displayed 100% Threat Confidence has been resolved.