Jun 13Vibe with Hermes Agent — Bengaluru · RSVP
ToolsMCPBlogResearchCommunityStar on GitHub
All Tools
A
OtherFreeOpen Source

AI INFRA GUARD

Full-stack AI red-teaming and security scanning platform

Apache-2.0

ABOUT

AI infrastructure introduces multiple new attack surfaces — model APIs, agent orchestration loops, tool-calling endpoints, and Model Context Protocol (MCP) servers — each requiring different scanning techniques. Traditional security tools lack AI-specific coverage. AI Infra Guard provides a unified platform that scans cloud-deployed model endpoints, tests agent logic for safety violations, probes MCP server implementations, and inventories AI skill dependencies. It gives security teams a single pane of glass for AI infrastructure risk across the full stack.

INTEGRATION GUIDE

1. Scan cloud-deployed LLM endpoints for prompt injection and data leakage vulnerabilities 2. Audit agent orchestration logic for safety boundary violations and tool misuse 3. Probe MCP server implementations for authentication and authorization weaknesses 4. Generate consolidated AI security reports for compliance and risk management

TAGS

securityred-teamingllmvulnerability-scanningmcpagent-security