ClawHub
Back to skill

Security audit

Lobster Autodream

Security checks for vulnerabilities and agentic risk

Overview

This skill is a disclosed local memory organizer; its main risk is that it may change long-term memory automatically.

Install this only if you want the agent to maintain local long-term memory from conversation history. Consider requiring a visible diff or confirmation before MEMORY.md is changed, and avoid storing secrets, sensitive personal data, or speculative inferences in memory.

SkillSpector

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

Vague Triggers

Medium
Confidence
90% confidence
Finding
The skill declares automatic triggering on broad system events such as heartbeat checks and session compression without defining strict activation boundaries, authorization checks, or user confirmation. In a memory-management skill that reads conversation history and updates persistent memory, ambiguous triggers can cause unintended execution and persistent data changes outside the user's clear intent.

Missing User Warnings

Medium
Confidence
96% confidence
Finding
The skill explicitly instructs the agent to update MEMORY.md automatically, but it does not require a user-facing warning or consent before modifying persistent storage. Because MEMORY.md is long-lived and may contain sensitive or user-specific information, silent writes can create privacy, integrity, and surprise-persistence risks, especially when triggered automatically during heartbeat or compression flows.

Static analysis

No suspicious patterns detected.