ClawHub
Back to skill

Security audit

Plea

Security checks across malware telemetry and agentic risk

Overview

This skill has no hidden executable code, but it tells agents to perform high-impact legal work automatically without clear user approval or limits.

Install only if you can constrain this to draft-only legal assistance. Do not allow it to scan broad private data, choose legal strategy, submit court filings, or respond to complaints unless a qualified human explicitly reviews and approves each action.

SkillSpector

By NVIDIA
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
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Findings (4)

Vague Triggers

Medium
Confidence
95% confidence
Finding
The manifest description is extremely broad and describes a system that could activate for many legal-adjacent tasks without clear scope, triggers, or user-consent boundaries. In an agent ecosystem, overbroad activation increases the chance the skill is invoked on sensitive matters and performs legal drafting or filing behavior in contexts the user did not explicitly authorize.

Missing User Warnings

High
Confidence
97% confidence
Finding
The skill explicitly advertises 'instant judicial submission' through e-filing integration but provides no warning about transmitting user data to courts or external filing systems, nor any requirement for explicit informed consent. That creates a serious risk of unauthorized legal action, disclosure of sensitive information, and irreversible real-world consequences from autonomous filings.

Missing User Warnings

Medium
Confidence
94% confidence
Finding
The fact-finding step says the agent scans logs, emails, and transaction hashes without any privacy notice, scope limitation, or consent model. Because these sources commonly contain highly sensitive personal, financial, and privileged information, the skill encourages broad data access beyond what may be necessary for a specific legal task.

Missing User Warnings

High
Confidence
98% confidence
Finding
Automatically generating and filing a response to a complaint to avoid default judgments is autonomous legal action with potentially binding consequences, yet the skill provides no warning, review gate, or requirement for user authorization. In the legal context, incorrect or unauthorized responses can waive defenses, create admissions, miss jurisdiction-specific requirements, or expose the user to major procedural harm.

VirusTotal

65/65 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.