ClawHub
Back to skill

Security audit

NeuralMind

Security checks across malware telemetry and agentic risk

Overview

This is a disclosed local code-retrieval skill with broad code-question triggers and persistent project memory that users should understand before installing.

Install this only for repositories where local indexing and persistent learned associations are acceptable. For sensitive codebases, confirm where .neuralmind data is stored, decide whether cross-agent memory sharing is acceptable, and use the documented environment toggles or project isolation if you do not want synapse recall/export behavior.

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 (5)

Vague Triggers

Medium
Confidence
91% confidence
Finding
The trigger list is very broad and composed of common developer phrases like 'how does', 'where is', and 'find function'. In a multi-skill environment, this can cause the skill to activate unexpectedly for ordinary conversation, steering the agent toward this skill's workflow and tool usage without clear user intent.

Missing User Warnings

Medium
Confidence
88% confidence
Finding
The skill describes persistent learned associations stored across sessions and potentially shared across multiple hosts, but does not present a clear user-facing privacy warning or consent requirement. This can expose sensitive repository structure, developer activity patterns, or cross-session context in ways users may not expect.

Shadow Command Trigger

Medium
Category
Trigger Abuse
Confidence
84% confidence
Finding
The trigger phrase 'find function' overlaps with a common built-in intent around searching or locating code. That creates a shadow-trigger risk where ordinary commands may invoke this skill unexpectedly, altering routing and potentially causing unintended tool access or context retrieval.

Shadow Command Trigger

Medium
Category
Trigger Abuse
Confidence
84% confidence
Finding
The trigger phrase 'find class' is similarly generic and likely to collide with ordinary code-navigation requests or host-native search behavior. In practice, this can cause the skill to preempt other safer or more appropriate handlers and broaden unintentional activation.

Shadow Command Trigger

Medium
Category
Trigger Abuse
Confidence
81% confidence
Finding
The phrase 'explain module' is a broad natural-language request that could match many benign coding conversations. Because the skill then encourages repository indexing and retrieval workflows, an accidental activation could expose more repository context than the user intended for a simple explanation request.

VirusTotal

64/64 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.