AI Sommelier — agentic threat model
The AI Sommelier is a low-to-moderate risk conversational agent focused on product recommendations. Its primary security exposure lies in its integration with live retail inventory databases and the potential for prompt injection to bypass age-gating or manipulate product suggestions.
OWASP AIVSS score rationale
| Autonomy of Action | 0.30 | |
| Goal-Driven Planning | 0.20 | |
| Self-Modification | 0.00 | |
| Dynamic Tool Use | 0.30 | |
| Persistent Memory | 0.10 | |
| Contextual Awareness | 0.40 | |
| Dynamic Identity | 0.10 | |
| Multi-Agent Interactions | 0.00 | |
| Non-Determinism | 0.50 | |
| Opacity & Reflexivity | 0.40 |
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.
Not certain from the listing — likely relies on a commercial foundation LLM to parse user preferences and generate conversational recommendations. Main threats include prompt injection to bypass age-gating or generate offensive content.
The agent connects to live retail inventory data to provide real-time recommendations. Threats include inventory database poisoning, unauthorized disclosure of proprietary stock/pricing data, and RAG extraction attacks.
Not certain from the listing — likely uses a basic RAG orchestration framework to translate user queries into inventory search parameters. Threats include insecure tool integration with the inventory API.
Not certain from the listing — deployed as a closed-source customer service widget or API. Threats include standard web application vulnerabilities, lack of API rate limiting, and insecure hosting infrastructure.
Not certain from the listing — no observability or guardrail mechanisms are mentioned. Threats include a lack of monitoring for hallucinated product details or malicious prompt injections.
Not certain from the listing — closed-source with no security certifications specified. Compliance risks are elevated due to the sale of age-restricted products (alcohol), requiring robust age-verification controls.
The agent operates as a standalone customer-facing tool with no described multi-agent or ecosystem marketplace interactions, minimizing horizontal cascading risks.
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.