Fraudlabs Pro

• Outcome: Review
• Latest audit: May 29, 2026 at 9:37 PM UTC
• Version:
  1.0.4

Overview

This FraudLabs Pro skill is transparent about its integration purpose, but it gives broad authenticated power to change fraud controls and business records without clear confirmation boundaries.

Install only if you are comfortable giving an agent access to your FraudLabs Pro account through Membrane. Use a least-privilege connection, verify or pin the Membrane CLI where possible, and require the agent to show exact parameters and get approval before creating, updating, deleting, blacklisting, whitelisting, sending SMS, or using raw proxy requests.

SkillSpector

By NVIDIA

Vulnerability Patterns

• Excessive Agency: Unrestricted Tool Access, Autonomous Decision Making, Scope Creep
• Prompt Injection: Instruction Override, Hidden Instructions, Exfiltration Commands
• Data Exfiltration: External Transmission, Env Variable Harvesting, File System Enumeration
• Privilege Escalation: Excessive Permissions, Sudo/Root Execution, Credential Access
• Supply Chain: Unpinned Dependencies, External Script Fetching, Obfuscated Code

Findings (1)

Missing User Warnings

Medium

• Confidence: 92% confidence
• Finding: The skill advertises destructive capabilities such as creating and deleting blacklist/whitelist entries and updating transaction records without clearly warning that these actions can alter fraud controls and business records. In an agent context, that increases the chance of unintended high-impact changes if a user request is ambiguous or the agent acts without explicit confirmation.

VirusTotal

65/65 vendors flagged this skill as clean.