ClawHub
Back to skill

Security audit

Email MCP Helper

Security checks across malware telemetry and agentic risk

Overview

This is a transparent email-tool reference, but it gives an agent broad multi-account email access and mutation power without enough built-in consent and safety boundaries.

Install only if you control and trust the MCP server and proxy, understand which accounts are configured, and can protect the API key and email credentials. Add separate agent rules requiring explicit user approval before sending, forwarding, replying, scheduling, deleting, moving, labeling, downloading attachments, or running bulk email operations.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • 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
93% confidence
Finding
This skill exposes broad read, search, send, delete, move, and attachment-download capabilities over potentially multiple email accounts, but it does not present clear user-facing privacy boundaries, consent expectations, retention limits, or warnings about the sensitivity of mailbox contents. In context, the skill is especially risky because it is a tool-reference skill that normalizes powerful email operations while delegating safe usage to another unspecified skill, increasing the chance that an agent or user enables invasive access without understanding the data-handling implications.

VirusTotal

61/61 vendors flagged this skill as clean.

View on VirusTotal