ClawHub
Back to skill

Security audit

brainstorming

Security checks for vulnerabilities and agentic risk

Overview

The skill is not overtly malicious, but its broad command execution access is poorly scoped for a vague brainstorming/data-processing purpose.

Review this skill carefully before installing. It has no embedded malware in the inspected artifact, but it asks for command execution under a vague automation purpose; only use it where shell access is sandboxed or explicitly approved per action.

A.I.G

Vulnerability Patterns
  • Skill Instruction HijackingAlters the agent's session goals or safety constraints when the skill loads
  • Agent Memory PoisoningWrites attacker-controlled rules into memory that affect later sessions
  • Remote Payload Retrieval and ExecutionFetches external code whose behavior can change after review
  • Embedded Malicious CodeShips malicious scripts inside the skill and executes them locally
  • Unauthorized Access and Privilege EscalationObtains permissions beyond the task's legitimate needs

SkillSpector

Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
Findings (4)

Context-Inappropriate Capability

High
Category
Not specified by scanner
Confidence
97% confidence
Finding

The skill declares generic exec access even though the documented purpose is brainstorming/data processing, which does not inherently require arbitrary shell execution. This unnecessarily expands the attack surface: prompt-influenced inputs, file paths, or operational instructions could be turned into command execution with host-level effects such as data exfiltration, file modification, or staging additional payloads.

Content

No source excerpt is available for this finding.

Vague Triggers

High
Category
Not specified by scanner
Confidence
96% confidence
Finding

The trigger condition ('when the user needs to perform related operations, automatically trigger') is overly broad and ambiguous, making it easy for the skill to activate in situations not clearly intended by the user. In a skill that also references command execution, file access, and external API usage, accidental invocation increases the chance of unintended sensitive actions or overreach.

Content

No source excerpt is available for this finding.

Intent-Code Divergence

Medium
Category
Not specified by scanner
Confidence
93% confidence
Finding

The skill advertises a safety control ('only execute whitelisted commands') while exposing a generic exec capability and providing no implementation or enforceable policy demonstrating that restriction. This creates a misleading trust boundary: downstream agents or users may rely on a non-existent guardrail and allow command execution in contexts where arbitrary shell access could occur.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
91% confidence
Finding

The documentation discusses API keys, external connections, file permissions, and command execution, but it does not present a clear user-facing warning that these operations may read/write files, invoke external services, or run commands. Without explicit disclosure and consent boundaries, users may provide sensitive inputs or approve use of the skill without understanding the operational risk.

Content

No source excerpt is available for this finding.

Static analysis

No suspicious patterns detected.