Rugproof — agentic threat model
Rugproof is a Web3 wallet and contract scanner with high-impact risk; while its autonomy is limited by the need for user transaction signatures, a compromise of its analysis engine or frontend could lead to devastating wallet-draining attacks via deceptive 'nuke' recommendations.
OWASP AIVSS score rationale
| Autonomy of Action | 0.30 | |
| Goal-Driven Planning | 0.20 | |
| Self-Modification | 0.10 | |
| Dynamic Tool Use | 0.50 | |
| Persistent Memory | 0.20 | |
| Contextual Awareness | 0.40 | |
| Dynamic Identity | 0.10 | |
| Multi-Agent Interactions | 0.20 | |
| Non-Determinism | 0.30 | |
| 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 — The specific LLMs used for contract analysis are undisclosed. Adversaries could craft obfuscated smart contracts (adversarial examples) specifically designed to bypass the AI's detection heuristics.
Not certain from the listing — The agent relies on real-time blockchain data and likely a database of known malicious signatures. Threats include poisoning of the threat intelligence database or manipulation of RPC nodes providing the blockchain state.
Not certain from the listing — The orchestration framework connecting the scanner to wallet APIs is proprietary. A key threat is prompt injection via malicious smart contract source code or metadata designed to hijack the agent's reasoning.
Not certain from the listing — Hosted deployment details are private. If the hosting infrastructure or frontend is compromised, attackers could swap legitimate transaction payloads with malicious ones, leading to direct theft of user funds.
Not certain from the listing — No observability or drift detection mechanisms are detailed. Gaps here could result in silent failures where new, complex rug-pull mechanisms (e.g., dynamic honeypots) go undetected.
Not certain from the listing — Compliance and security standards are not specified. The primary risk is the lack of decentralized or multi-signature validation on the threat definitions, creating a single point of failure.
Not certain from the listing — While part of the ForgeX ecosystem, direct multi-agent interactions are not detailed. If integrated with ForgeX's governance trading bot, a compromised scanner could feed false threat data to trigger automated panic-selling.
MAESTRO — the 7-layer agentic threat-modeling framework (Cloud Security Alliance / Ken Huang).
These scores are auto-generated from public information (the agent's own listing, docs, and repository) using the canonical OWASP AIVSS formula and the MAESTRO framework — an estimate for guidance, not a penetration test, audit, or certification. See the scoring methodology. Are you the vendor? Factual corrections are free.