fix: branding change from javelin to highflame

[abhijitjavelin]

Description:

• Add Highflame guardrails plugin for AI security with trust & safety and prompt injection detection
• Support multi-guardrail endpoint (/v1/guardrails/apply) with configurable guardrails
• Default guardrails: trustsafety (threshold 0.75) and promptinjectiondetection (threshold 0.8)
• Support for additional guardrails: lang_detector, dlp_gcp, model_armor
• Configurable per-guardrail thresholds via parameters
• Works with both beforeRequestHook (input) and afterRequestHook (output)

Tests Run/Test cases added:

• Basic request passes guardrails (safe content)
• Malicious content blocked by trustsafety (weapons detection)
• Prompt injection detected and blocked (96% confidence)
• Output guardrails validate LLM responses

Type of Change:

• Bug fix (non-breaking change which fixes an issue)
• New feature (non-breaking change which adds functionality)
• Breaking change (fix or feature that would cause existing functionality to not work as expected)
• Documentation update
• Refactoring (no functional changes)