ClawHub
Back to skill

Security audit

Email Writer

Security checks across malware telemetry and agentic risk

Overview

This email skill is transparent about using your logged-in mailbox, but it can click Send in real webmail sessions without a required final approval step.

Install only if you are comfortable letting an agent operate your logged-in webmail session. Use it for drafting first, review the account, recipient, subject, and body yourself, and require an explicit confirmation before any send action.

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
95% confidence
Finding
The trigger phrases are broad enough to activate on ordinary requests like drafting or sending email, which can cause the skill to engage browser automation on a live mailbox with minimal friction. In this skill’s context, that is materially risky because activation can lead directly to reading message context and sending email on the user’s behalf through an authenticated browser session.

VirusTotal

66/66 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.