Marb.ai — agentic threat model
Marb.ai presents a moderate-to-high risk profile due to its integration with sensitive CRM data, automated email/chat processing, and capability to draft financial offers, making it highly susceptible to indirect prompt injection and data exfiltration.
OWASP AIVSS score rationale
| Autonomy of Action | 0.70 | |
| Goal-Driven Planning | 0.60 | |
| Self-Modification | 0.10 | |
| Dynamic Tool Use | 0.50 | |
| Persistent Memory | 0.60 | |
| Contextual Awareness | 0.80 | |
| Dynamic Identity | 0.20 | |
| Multi-Agent Interactions | 0.50 | |
| Non-Determinism | 0.70 | |
| Opacity & Reflexivity | 0.60 |
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 utilizes commercial LLMs combined with a proprietary speech-to-text model fine-tuned on real estate jargon. Primary threats include prompt injection via incoming emails/chats and model hallucinations when drafting property offers.
Utilizes vector-based similarity search across a global property database and processes buyer profiles. Key threats include vector database poisoning, embedding inversion, and unauthorized exfiltration of sensitive buyer PII.
Orchestrates automated research, task extraction from voice/emails, and CRM API integration. Vulnerable to indirect prompt injection where malicious text in emails or chats triggers unauthorized CRM tool execution or data manipulation.
Not certain from the listing — likely hosted on cloud infrastructure to support 24/7 background research and API integrations. Risks include insecure storage of CRM API keys and lack of sandboxing when processing untrusted voice/text inputs.
Not certain from the listing — no mention of continuous monitoring, guardrails, or drift detection. Risks include undetected drift in property matching algorithms or unmonitored generation of inappropriate/unauthorized follow-up communications.
Not certain from the listing — no explicit compliance certifications (e.g., SOC 2, GDPR) or fine-grained RBAC mentioned. Risks include unauthorized access to the CRM integration and non-compliance with data privacy regulations regarding buyer PII.
Employs multiple AI agents for 24/7 research and integrates with external CRM ecosystems. Risks include cascading failures if the CRM API changes, and trust abuse between the research agents and the main assistant agent.
MAESTRO — the 7-layer agentic threat-modeling framework (Cloud Security Alliance / Ken Huang).