🔍 Agentic AI Vulnerability Scanner

Proactive Security Assessment Against 7 Critical Vulnerability Categories

💰 Vulnerability Scan: 350 Gold Coins 🪙

🤖 AI Agent Identity

Agent / Project Name *

Agent Description *

Agent Framework *

LLM Provider *

⚙️ Tools & Capabilities

Select all tools/capabilities your agent uses:

Web Browsing / URL Fetching Code Execution (Python, JS, etc.) File Read / Write Operations Database Queries (SQL, NoSQL) External API Calls Email Sending / Messaging Shell / OS Commands MCP Servers / Tool Servers Persistent Memory / State Image Generation

🔐 Authentication & Access Control

Authentication Mechanism *

None API Key Basic Auth JWT Token OAuth 2.0 Mutual TLS Custom Token

Access Control Model *

None Basic (Admin/User) RBAC ABAC

🔑 Token Handling & Input Security

How are tokens/credentials handled? *

Plaintext / Config Files Environment Variables Secrets Manager / Vault Encrypted Storage

Input Sanitization Level *

None Basic Filtering ML-Based Detection Instruction Hierarchy

📦 Supply Chain & Deployment

Dependency Management *

No Pinning Version Pinned Hash Pinned Private Registry

Deployment Type *

Public Cloud Private Cloud On-Premise Hybrid

Data Flow Description

🛡️ Security Features

Uses MCP Servers

Multi-Agent Architecture

Human-in-the-Loop

Logging Enabled

Sandboxed Execution

Rate Limiting Enabled

💡 Note: Your architecture details are processed securely to generate a vulnerability assessment against the 7 core Agentic AI vulnerability categories: Token Theft, Token Passthrough, Rug Pull Attack, Prompt Injection, Command Injection, Tool Poisoning, and Unauthenticated Access.