Skip to main content

Features

KoreShield provides a comprehensive set of security capabilities for protecting LLM integrations.

Core Security​

  • Attack Detection - Multi-layered prompt injection detection with keyword rules, pattern analysis, custom DSL, and ML heuristics
  • RAG Defense Engine - Scan retrieved documents for indirect prompt injection before they reach your LLM
  • Security Overview - How sanitization, detection, and policy enforcement work together

Platform Features​

  • RBAC - Role-based access control for team access and dashboard permissions
  • Monitoring & Alerting - Prometheus metrics, rule-based alerting, and multi-channel notifications (Email, Slack, Teams, PagerDuty)