Drift

Security checks across malware telemetry and agentic risk

Overview

This Drift integration is coherent, but it can modify or delete CRM records through Membrane without clear in-skill safeguards for high-impact actions.

Install only if you trust Membrane with the intended Drift workspace. Before using it, require the agent to fetch and show the exact record, confirm the specific action and ID, and get explicit approval before any create, update, delete, message-send, or raw proxy request. Consider pinning or reviewing the Membrane CLI package version before global installation.

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
92% confidence
Finding
The skill advertises destructive actions such as deleting accounts and contacts, but provides no guidance to require explicit user confirmation, verify target identity, or use safeguards before execution. In an agentic context, this increases the risk of accidental or overly broad destructive operations caused by ambiguous prompts, hallucinated identifiers, or unsafe automation.

VirusTotal

65/65 vendors flagged this skill as clean.

View on VirusTotal