Fabi.ai — agentic threat model
Fabi.ai presents a moderate-to-high risk profile primarily due to its direct integration with enterprise data sources, where prompt injection could lead to unauthorized data extraction or SQL execution.
OWASP AIVSS score rationale
| Autonomy of Action | 0.40 | |
| Goal-Driven Planning | 0.50 | |
| Self-Modification | 0.10 | |
| Dynamic Tool Use | 0.70 | |
| Persistent Memory | 0.30 | |
| Contextual Awareness | 0.50 | |
| Dynamic Identity | 0.20 | |
| Multi-Agent Interactions | 0.10 | |
| Non-Determinism | 0.50 | |
| Opacity & Reflexivity | 0.40 |
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 third-party foundation models to translate natural language into SQL/code. Key threats include prompt injection that bypasses system instructions to access unauthorized data.
Utilizes direct data source connectors to query database schemas and records. This creates a high risk of data exfiltration and exposure of sensitive business intelligence if the agent is compromised.
Orchestrates workflows and executes queries against connected databases. Vulnerabilities include insecure tool integration, where the agent might execute destructive write queries or be manipulated into SQL injection.
Not certain from the listing — likely hosted as a closed-source SaaS. Main threats involve the secure storage of database credentials/secrets and potential lateral movement if the hosting environment is breached.
Not certain from the listing — requires robust logging of executed queries and LLM inputs. Gaps here could lead to undetected data harvesting or silent prompt injection attacks.
Not certain from the listing — requires strict tenant isolation and credential encryption. Lack of fine-grained access controls could allow users to query data sources they are not authorized to see.
Not certain from the listing — operates primarily as a standalone data assistant. Ecosystem risks are minimal unless it integrates with external third-party agent marketplaces.
MAESTRO — the 7-layer agentic threat-modeling framework (Cloud Security Alliance / Ken Huang).