AgentReadyHomeAgent ListingRuntimePricing
One free check per registered account

Check your AI agent's security readiness

Answer a few questions about your AI agent. You'll get an instant readiness score mapped to NIST AI RMF, the OWASP LLM Top 10, and ISO/IEC 42001 — then can unlock a full report with prioritized, cited fixes as an actionable checklist.

Your first snapshot is free. Each registered account gets one free check. The full fix-list report is $129 (Indie) · $499 (Team) · $1,999 (Enterprise) — see what's included, or view a sample report →

This is an AI-generated readiness assessment for guidance only — not a certification, audit, or penetration test.

Describe your AI agent

Tell us what it does, what data and tools it can access, how independently it acts, whether it has memory or talks to other agents, and what approval or security controls are in place.

Use either this free-form description or the security questions below. Both paths produce a MAESTRO 7-layer review and an AIVSS-informed risk estimate.

Your AI agent

A little context so the grade fits your deployment.

1Where is your AI agent deployed?
2Most sensitive data it can access?

Governance — NIST AI RMF

How your organization governs AI risk.

3Do you keep an inventory of your AI systems / agents?
4Is there a named owner accountable for AI risk?
5Do you have an acceptable-use / AI policy?
6How prepared is your AI incident-response plan?

LLM & agent security — OWASP LLM Top 10

The controls attackers probe first on tool-using agents.

7How strong are your prompt-injection defenses?
8How thoroughly do you validate / sanitize model output before acting on it?
9Do you rate-limit and cap spend on agent / model calls?
10How well do you prevent sensitive-data leakage in responses?
11How tightly is the agent's tool access / autonomy scoped (least privilege)?
12How robust is human oversight for high-impact / irreversible actions?

Management — ISO/IEC 42001

Whether security is operationalized, not one-off.

13Is your AI management approach documented?
14How comprehensive is your logging / monitoring of agent decisions?
15How mature is your review / continual-improvement cycle?

Agentic AI security — OWASP Agentic Top 10 & MAESTRO

Risks that only emerge when the agent plans, holds memory, calls tools, and coordinates. (Skip if not an autonomous agent.)

16How strong are your defenses against agent goal hijack (injected goals / plans)?
17Does the agent have a distinct identity with scoped, short-lived credentials?
18How well do you vet the agent's tools, skills, and MCP servers (agentic supply chain)?
19How well do you protect the agent's memory / context from poisoning?
20How secure is inter-agent communication? (single-agent → answer Strong/Full)
21How traceable are the agent's decisions and tool calls (logging / audit)?

Used only for your receipt if you purchase. Your report opens on the next screen — bookmark it to return.

Your answers are encrypted at rest, used only to generate your report, and never used to train models.

Enter at least 20 characters about your agent, or answer 4 security questions below.