AgentReadyHomeAgent ListingPricing

← AI Refinery

AI Refinery — agentic threat model

8.1AIVSS 8.1 · High

AI Refinery is a high-risk enterprise orchestration platform that coordinates multi-agent teams across siloed business systems, presenting a significant attack surface if agent-to-agent trust or tool integrations are compromised.

OWASP AIVSS score rationale

AIVSS = (CVSS_Base + AARS) × Mitigation_Factor, where AARS = (10 − CVSS_Base) × (Factor_Sum / 10) × ThM
CVSS base 8.5AARS uplift 1.02Factor sum 6.5/10Threat ×1.05Mitigation ×0.85
Autonomy of Action
0.70
Goal-Driven Planning
0.80
Self-Modification
0.30
Dynamic Tool Use
0.80
Persistent Memory
0.60
Contextual Awareness
0.80
Dynamic Identity
0.50
Multi-Agent Interactions
0.90
Non-Determinism
0.60
Opacity & Reflexivity
0.50

Scored with the canonical OWASP AIVSS formula (AIVSS calculator reference); agentic risk factors estimated from the agent’s described capabilities.

MAESTRO 7-layer threat model

Per-layer threats for this agent. Layers tagged “not certain from listing” are general, caveated commentary where the public description didn’t pin that layer.

L1 · Foundation Models✓ mapped

Utilizes NVIDIA AI Enterprise foundation models and NIMs. Threats include adversarial prompt injection bypassing guardrails and model reprogramming within the enterprise context.

L2 · Data Operations⚠ not certain from listing

Not certain from the listing — The platform connects siloed AI technologies and embeds prebuilt process knowledge, but specific vector database integrations, RAG architectures, or data lineage controls are not detailed.

L3 · Agent Frameworks✓ mapped

Supports building and adapting AI agent teams. Threats include insecure tool integration, cascading failures across agent workflows, and malicious tool manipulation by compromised agents.

L4 · Deployment & Infrastructure✓ mapped

Built on NVIDIA AI Enterprise, implying containerized microservices. Threats include container escape, insecure API endpoints connecting siloed systems, and privilege escalation within the hosting environment.

L5 · Evaluation & Observability✓ mapped

Explicitly manages AI components with a focus on cost, accuracy, security, and responsible use, indicating built-in evaluation and guardrail mechanisms, though specific logging and drift detection capabilities are not detailed.

L6 · Security & Compliance (cross-cutting)✓ mapped

Designed for enterprise deployment with a focus on security and responsible use. However, specific compliance certifications (e.g., SOC2, ISO) or identity and access management (IAM) integrations are not explicitly detailed.

L7 · Agent Ecosystem✓ mapped

Enables the creation of collaborative AI agent teams. This introduces significant risks of agent-to-agent trust abuse, lateral movement of malicious instructions between agents, and cascading failures across the ecosystem.

MAESTRO — the 7-layer agentic threat-modeling framework (Cloud Security Alliance / Ken Huang).

These scores are auto-generated from public information (the agent's own listing, docs, and repository) using the canonical OWASP AIVSS formula and the MAESTRO framework — an estimate for guidance, not a penetration test, audit, or certification. See the scoring methodology. Are you the vendor? Factual corrections are free.