AgentReadyHomeAgent Listing

← TaxGPT

TaxGPT — agentic threat model

8.5AIVSS 8.5 · High

TaxGPT handles highly sensitive financial data and PII, presenting a high-impact target for data exfiltration and tax fraud, though its operational risk is moderated by its likely deployment as a human-in-the-loop advisory tool.

OWASP AIVSS score rationale

AIVSS = (CVSS_Base + AARS) × Mitigation_Factor, where AARS = (10 − CVSS_Base) × (Factor_Sum / 10) × ThM
CVSS base 7.5AARS uplift 0.98Factor sum 3.9/10Threat ×1.0Mitigation ×1.0
Autonomy of Action
0.40
Goal-Driven Planning
0.50
Self-Modification
0.10
Dynamic Tool Use
0.40
Persistent Memory
0.50
Contextual Awareness
0.70
Dynamic Identity
0.20
Multi-Agent Interactions
0.10
Non-Determinism
0.50
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 — likely relies on third-party frontier LLMs optimized for tax domain tasks. Primary threats include prompt injection leading to hallucinated tax advice or leaking of system prompts.

L2 · Data Operations⚠ not certain from listing

Not certain from the listing — likely utilizes RAG over tax codes, multi-jurisdictional regulations, and uploaded client documents. Vulnerable to knowledge-base poisoning if tax updates are manipulated, and data exfiltration of sensitive financial PII.

L3 · Agent Frameworks⚠ not certain from listing

Not certain from the listing — orchestrates document analysis, memo generation, and email drafting. Vulnerable to insecure tool execution if the agent directly interfaces with email clients or document management systems.

L4 · Deployment & Infrastructure⚠ not certain from listing

Not certain from the listing — hosted as a closed-source SaaS platform. Threats include inadequate sandboxing of uploaded tax documents (e.g., malicious PDFs) and insecure storage of sensitive financial records.

L5 · Evaluation & Observability⚠ not certain from listing

Not certain from the listing — no details on observability or guardrails. Gaps here could allow hallucinated tax advice or outdated regulatory interpretations to go unnoticed, leading to severe compliance violations.

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

Not certain from the listing — handling tax and financial data requires strict compliance (e.g., GLBA, SOC2, GDPR). The listing does not specify access controls, encryption standards, or audit logging mechanisms.

L7 · Agent Ecosystem⚠ not certain from listing

Not certain from the listing — operates primarily as a standalone digital worker. Ecosystem threats are low unless it integrates directly with external accounting platforms or IRS e-file APIs.

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