AI EdWiBo — agentic threat model
AI EdWiBo is a low-autonomy educational assistant focused on grading and feedback, presenting low systemic risk but requiring strong data privacy controls to protect student PII and ensure grading integrity.
OWASP AIVSS score rationale
| Autonomy of Action | 0.20 | |
| Goal-Driven Planning | 0.10 | |
| Self-Modification | 0.00 | |
| Dynamic Tool Use | 0.10 | |
| Persistent Memory | 0.20 | |
| 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 relies on commercial LLMs prompted with specific exam board criteria. Threats include prompt injection from student essays designed to force a high grade or bypass evaluation logic.
Not certain from the listing — processes student essays and stores exam criteria. Threats include data exfiltration of student submissions (which may contain PII) and potential poisoning of the reference criteria.
Not certain from the listing — orchestration likely manages the flow from student input to exam-aligned feedback. Threats include insecure handling of untrusted student text within the prompt template.
Not certain from the listing — hosted as a web application with a dashboard. Threats include standard web application vulnerabilities (OWASP Top 10) and unauthorized access to the teacher/institution dashboard.
Not certain from the listing — requires robust monitoring to ensure grading consistency and prevent drift in assessment standards over time.
Not certain from the listing — handling student data requires strict compliance with privacy regulations (such as GDPR, COPPA, or FERPA), but no specific compliance certifications are cited.
Not certain from the listing — appears to operate as a standalone tool with no multi-agent or marketplace interactions described.
MAESTRO — the 7-layer agentic threat-modeling framework (Cloud Security Alliance / Ken Huang).