Braintree

Security checks across malware telemetry and agentic risk

Overview

This Braintree skill appears legitimate, but it gives an agent access to sensitive payment and customer-changing actions without clear approval safeguards.

Install only if you intentionally want an agent to operate on your Braintree account through Membrane. Use sandbox or least-privileged access where possible, verify the CLI package, and require human approval before any charge, refund, void, capture, dispute, vaulting, or delete action.

SkillSpector

By NVIDIA

Vulnerability Patterns

Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep

Findings (1)

Missing User Warnings

Medium

Confidence: 87% confidence
Finding: The skill advertises destructive and financially sensitive operations such as deleting customers/payment methods, voiding transactions, refunding transactions, and accepting disputes without any warning, confirmation guidance, or approval boundary. In an agent setting, this increases the chance that a model may invoke irreversible or high-impact actions on ambiguous prompts, causing financial loss, data loss, or business workflow disruption.

VirusTotal

65/65 vendors flagged this skill as clean.

View on VirusTotal