ClawHub
Back to skill

Security audit

brainstorming

Security checks across malware telemetry and agentic risk

Overview

This skill is very broad and asks for command execution despite only describing a vague brainstorming and automation purpose.

Review before installing. This does not show clear malicious behavior, but it gives an agent broad command and file capability under a vague automation description. Only use it in a constrained environment and avoid granting real credentials or sensitive workspace access unless the publisher narrows the scope and documents exact allowed operations.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • 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
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
Findings (3)

Context-Inappropriate Capability

Medium
Confidence
92% confidence
Finding
The skill declares the exec tool even though the stated purpose is generic brainstorming/data processing, creating an unnecessary command-execution capability gap. In an agent setting, this broadens the attack surface significantly because prompt injection or ambiguous task routing could cause arbitrary local commands to be run without a narrowly justified business need.

Description-Behavior Mismatch

Medium
Confidence
89% confidence
Finding
The manifest presents the skill as a generic brainstorming/automation tool, while the body describes API access, file operations, network use, and command execution that materially exceed that framing. This mismatch is dangerous because reviewers, orchestrators, or users may trust the skill in low-risk contexts while it quietly requests capabilities suitable for far more invasive actions.

Vague Triggers

Medium
Confidence
84% confidence
Finding
The trigger condition states the skill activates whenever a user needs related operations, which is too broad to safely constrain invocation. In practice, vague triggers increase the chance the skill is selected in unintended contexts, where its declared read/exec capabilities could be applied to requests that never warranted such access.

VirusTotal

VirusTotal findings are pending for this skill version.

View on VirusTotal

Static analysis

No suspicious patterns detected.