wshobson/agents — agentic threat model
This agent marketplace presents a high-risk profile due to its integration with Claude Code, granting 80+ specialized subagents potential access to local developer environments, filesystems, and terminal execution without explicit sandboxing or verification controls.
OWASP AIVSS score rationale
| Autonomy of Action | 0.70 | |
| Goal-Driven Planning | 0.80 | |
| Self-Modification | 0.20 | |
| Dynamic Tool Use | 0.80 | |
| Persistent Memory | 0.30 | |
| Contextual Awareness | 0.60 | |
| Dynamic Identity | 0.10 | |
| Multi-Agent Interactions | 0.80 | |
| Non-Determinism | 0.70 | |
| 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.
Not certain from the listing — Assumes Claude models as the foundation. Primary threats include prompt injection leading to malicious code generation or unauthorized tool execution.
Not certain from the listing — Operates directly on local developer codebases and files. Threats include unauthorized reading of sensitive local files, exfiltration of hardcoded secrets, and codebase poisoning.
Utilizes Claude Code and custom marketplace definitions. High risk of tool misuse and insecure tool integration, as development subagents typically require powerful filesystem and terminal execution capabilities.
Not certain from the listing — Likely runs locally on developer workstations. Threats include local host compromise, privilege escalation, and lateral movement within the developer's network if a subagent is compromised.
Not certain from the listing — No built-in guardrails, logging, or observability mechanisms are mentioned for monitoring the execution of the 80+ subagents.
Not certain from the listing — Being a free, open-source repository, there are no mentioned security audits, compliance certifications, or access control policies governing the subagents.
High exposure. As a marketplace of 80+ subagents, it is highly vulnerable to supply chain attacks, malicious subagent definitions, and cascading failures when multiple subagents interact or orchestrate workflows.
MAESTRO — the 7-layer agentic threat-modeling framework (Cloud Security Alliance / Ken Huang).