Pilot Fraud Detection Pipeline Setup

• Outcome: Review
• Latest audit: May 29, 2026 at 9:35 PM UTC
• Version:
  1.0.0

Overview

The skill appears to be a legitimate fraud-operations workflow, but it can trigger account freezes and transaction declines without clearly documented safety gates.

Review this skill carefully before installing. Use it only in a sandbox or dry-run environment until you have confirmed exactly which downstream systems receive its fraud verdicts and blocked-entity events, and require human approval before enabling live account freezes or transaction declines.

SkillSpector

By NVIDIA

Vulnerability Patterns

• Data Exfiltration: External Transmission, Env Variable Harvesting, File System Enumeration
• Trigger Abuse: Overly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
• Prompt Injection: Instruction Override, Hidden Instructions, Exfiltration Commands
• Privilege Escalation: Excessive Permissions, Sudo/Root Execution, Credential Access
• Supply Chain: Unpinned Dependencies, External Script Fetching, Obfuscated Code

Findings (3)

Missing User Warnings

Medium

• Confidence: 92% confidence
• Finding: The README explicitly describes enforcement actions such as freezing accounts and declining transactions without any warning, approval gate, or guidance on safe testing. In a fraud pipeline, these actions can directly affect real users and funds, so normalizing them as routine setup behavior increases the chance of accidental harmful deployment or misuse.

Missing User Warnings

Medium

• Confidence: 95% confidence
• Finding: The example workflow instructs users to publish a confirmed fraud verdict and blocked-entity action as part of a demo, but does not warn that these messages may trigger real blocking, reversals, or account restrictions in connected systems. Because the skill is specifically for a live fraud enforcement pipeline, these examples materially increase the risk of operators executing customer-impacting actions during setup or testing.

Missing User Warnings

Medium

• Confidence: 90% confidence
• Finding: The enforcer role is explicitly designed to trigger external actions such as freezing accounts and declining transactions, but the skill does not present this with a prominent safety warning or require an explicit confirmation gate. In a fraud operations context this is legitimate functionality, but insufficient warning increases the risk of accidental disruptive actions against real users or systems.

VirusTotal

64/64 vendors flagged this skill as clean.