Trustcheck AI — agentic threat model
Trustcheck AI is a low-autonomy, passive analysis tool designed to assist human decision-making, presenting low agentic risk but high privacy sensitivity due to the processing of user-uploaded screenshots and messages.
OWASP AIVSS score rationale
| Autonomy of Action | 0.10 | |
| Goal-Driven Planning | 0.10 | |
| Self-Modification | 0.00 | |
| Dynamic Tool Use | 0.20 | |
| Persistent Memory | 0.10 | |
| Contextual Awareness | 0.30 | |
| Dynamic Identity | 0.00 | |
| Multi-Agent Interactions | 0.00 | |
| Non-Determinism | 0.40 | |
| Opacity & Reflexivity | 0.30 |
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 uses multimodal foundation models (OCR, vision, NLP) to analyze text, images, and screenshots. Vulnerable to adversarial inputs (e.g., prompt injection embedded in screenshots or text) designed to bypass scam detection.
Not certain from the listing — requires a database of known scam signatures, phishing URLs, and deepfake patterns. Vulnerable to poisoning of this reference database or evasion via zero-day scam variants.
Not certain from the listing — orchestration seems limited to single-turn analysis (ingest -> analyze -> output risk rating). Low risk of tool misuse, but parsing logic for URLs and images must be secure against buffer overflows or SSRF.
Not certain from the listing — deployed as a mobile app (iOS/Android) communicating with a backend API. Threats include API abuse, reverse engineering of the mobile client, and insecure transit of user-uploaded sensitive screenshots.
Not certain from the listing — requires continuous monitoring for drift in scam tactics and false positive/negative rates. Lack of transparent observability could allow silent failures where new scam techniques bypass detection.
Not certain from the listing — handles potentially sensitive user data (screenshots of private messages). Compliance risks include privacy violations (GDPR/CCPA) if uploaded images contain PII and are stored or used for model training without consent.
The agent operates as a standalone, closed-loop utility with no described multi-agent or marketplace integrations, minimizing ecosystem-level 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.