PickMyCall — agentic threat model
PickMyCall is a high-autonomy voice agent interacting directly with the public and integrating with SMB CRMs and scheduling tools, presenting risks of voice-based prompt injection, PII exposure, and unauthorized API actions in connected business systems.
OWASP AIVSS score rationale
| Autonomy of Action | 0.80 | |
| Goal-Driven Planning | 0.50 | |
| Self-Modification | 0.00 | |
| Dynamic Tool Use | 0.60 | |
| Persistent Memory | 0.40 | |
| Contextual Awareness | 0.50 | |
| Dynamic Identity | 0.10 | |
| Multi-Agent Interactions | 0.10 | |
| 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 proprietary or third-party LLMs coupled with STT/TTS models. Key threats include voice-based prompt injection (where a caller manipulates the LLM via spoken instructions) and conversational hallucinations leading to incorrect business commitments.
Not certain from the listing — processes real-time voice streams and customer contact details to sync with SMB databases. Threats include unauthorized retention or exposure of caller PII, and lack of clear data boundaries between different SMB tenants.
Not certain from the listing — orchestrates call flows, intent parsing, and tool execution (scheduling/CRM updates). Threats include insecure tool integration where malicious caller inputs trigger unintended API calls to connected SMB tools.
Not certain from the listing — relies on telephony/VoIP infrastructure and cloud hosting for the dashboard. Threats include toll fraud, SIP trunk exploitation, and unauthorized access to the mobile dashboard APIs.
Not certain from the listing — provides a mobile dashboard for business productivity, but lacks explicit details on real-time guardrails, call monitoring, or anomaly detection to flag abusive callers.
Not certain from the listing — handles customer PII and schedules appointments, necessitating compliance with regulations like GDPR, CCPA, and telephony-specific laws (e.g., TCPA), though no compliance certifications are listed.
Not certain from the listing — designed to be embedded seamlessly into third-party SMB partner solutions. This horizontal integration model introduces risks of cascading trust failures if a partner platform is compromised.
MAESTRO — the 7-layer agentic threat-modeling framework (Cloud Security Alliance / Ken Huang).