AgentReadyHomeAgent Listing

← Droxy AI

Droxy AI — agentic threat model

7.6AIVSS 7.6 · High

Droxy AI is a RAG-driven chatbot platform with moderate risk, primarily stemming from potential knowledge-base poisoning via user-uploaded content and prompt injection risks leading to data exfiltration or unauthorized actions on integrated platforms like Discord.

OWASP AIVSS score rationale

AIVSS = (CVSS_Base + AARS) × Mitigation_Factor, where AARS = (10 − CVSS_Base) × (Factor_Sum / 10) × ThM
CVSS base 6.5AARS uplift 1.12Factor sum 3.2/10Threat ×1.0Mitigation ×1.0
Autonomy of Action
0.30
Goal-Driven Planning
0.20
Self-Modification
0.10
Dynamic Tool Use
0.30
Persistent Memory
0.30
Contextual Awareness
0.60
Dynamic Identity
0.20
Multi-Agent Interactions
0.10
Non-Determinism
0.60
Opacity & Reflexivity
0.50

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.

L1 · Foundation Models⚠ not certain from listing

Not certain from the listing — The platform uses 'ChatGPT-powered' models. Threats include prompt injection, adversarial manipulation, and misaligned outputs, but the specific model versions and alignment guardrails are not disclosed.

L2 · Data Operations✓ mapped

Users upload diverse resources (PDFs, videos, website content) to build the chatbot's knowledge base. This introduces significant risks of data/knowledge-base poisoning, where malicious or manipulated files are uploaded to alter the bot's behavior or exfiltrate sensitive data via indirect prompt injection.

L3 · Agent Frameworks✓ mapped

The platform orchestrates RAG and persona settings. Vulnerabilities include insecure integration of the RAG pipeline, potential prompt leakage, and manipulation of the chatbot's persona/instructions through user inputs.

L4 · Deployment & Infrastructure⚠ not certain from listing

Not certain from the listing — Droxy AI is a hosted SaaS platform. Risks include insecure storage of uploaded files (PDFs, videos), lack of sandboxing during document parsing/extraction, and potential container or host compromise.

L5 · Evaluation & Observability⚠ not certain from listing

Not certain from the listing — There is no mention of built-in guardrails, interaction logging, anomaly detection, or evaluation frameworks to monitor chatbot behavior and detect malicious inputs.

L6 · Security & Compliance (cross-cutting)⚠ not certain from listing

Not certain from the listing — The listing does not specify compliance standards (e.g., GDPR, SOC2), access control mechanisms for uploaded data, or authentication protocols for managing the chatbot configurations.

L7 · Agent Ecosystem✓ mapped

The platform integrates with external ecosystems like Discord and websites. A compromised or manipulated chatbot could be used to distribute spam, launch social engineering attacks, or interact maliciously with other users and bots within those ecosystems.

MAESTRO — the 7-layer agentic threat-modeling framework (Cloud Security Alliance / Ken Huang).