Email Assistant

Security checks across malware telemetry and agentic risk

Overview

This is a simple email-help skill that works on user-provided text and does not request mailbox access, credentials, code execution, or automatic sending.

Install only if you are comfortable sharing selected email contents with the agent. Do not paste confidential messages unnecessarily, and review any drafted replies or extracted tasks before sending or acting on them.

SkillSpector

By NVIDIA

Vulnerability Patterns

Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
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

Findings (1)

Vague Triggers

Medium

Confidence: 91% confidence
Finding: The trigger examples are broad, natural-language phrases like asking for help processing an email or listing today's actionable items. Because they closely match ordinary conversation, the skill may activate in situations the user did not explicitly intend, causing unrelated or sensitive email-like text to be processed under this workflow. In an email assistant context, that increases the risk of over-collection or inappropriate handling of sensitive business communications.

VirusTotal

64/64 vendors flagged this skill as clean.

View on VirusTotal