feat: add findings engine and assessment reporting

[APonder-Dev]

Summary

This pull request introduces the initial findings and assessment engine for SentinelAudit.

SentinelAudit now performs basic interpretation of collected telemetry instead of functioning solely as a raw telemetry collector.

Major Changes

Added Findings Engine

• Added sentinel_audit/findings/
• Added firewall_findings.py
• Added structured finding generation

Added Security Analysis

• Added firewall telemetry analysis
• Added severity-based findings
• Added defensive recommendations
• Added informational, medium, and high severity classifications

Reporting Improvements

• Added Security Findings section to Markdown reports
• Added findings serialization to JSON output
• Improved assessment-oriented reporting structure

Testing Improvements

• Added findings engine unit tests
• Expanded pytest validation coverage
• Verified GitHub Actions CI compatibility

Validation Performed

• Ruff linting passed
• Pytest validation passed
• Windows execution validated
• Fedora Linux execution validated
• Markdown and JSON reporting verified

Example Finding Output

{
  "severity": "informational",
  "title": "Firewall status collected",
  "description": "Firewall telemetry was collected successfully.",
  "recommendation": "Review collected firewall details for profile-specific configuration issues."
}

Architectural Impact

This update transitions SentinelAudit from a telemetry collection utility into the foundation of a lightweight security assessment framework.