ClawHub
Back to skill

Security audit

smartclaws-master-agent

Security checks for vulnerabilities and agentic risk

Overview

This is a clearly described SmartClaws control procedure that may command configured devices and write on-chain, but its authority is disclosed and bounded by owner configuration.

Install this only in a SmartClaws workspace you control, with a reviewed SMARTCLAWS.md goal and narrow AGENTS.md permissions. Expect it to read configured telemetry, possibly disclose encrypted channel contents to the agent, spend wallet funds for SmartClaws writes, create permanent on-chain logs, notify configured agents, and command only devices marked commandable.

SkillSpector

Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (2)

Vague Triggers

Medium
Confidence
91% confidence
Finding
The trigger text is broad enough to match routine requests like checking devices, deciding what to do, commanding a device, or auditing decisions. In a skill that can publish on-chain commands to external devices, overbroad activation increases the chance the agent invokes this procedure in situations where the user intended discussion or analysis only, leading to unintended writes or device actions.

Missing User Warnings

Medium
Confidence
88% confidence
Finding
The skill performs potentially irreversible on-chain publishes and may command external devices, but the top-level description does not present a clear warning about these real-world side effects. Without an upfront warning, users may invoke the skill expecting analysis or monitoring and instead cause blockchain transactions or physical device state changes.

Static analysis

No suspicious patterns detected.