ClawHub
Back to skill

Security audit

Lumbox

Security checks across malware telemetry and agentic risk

Overview

The skill is transparent about its purpose, but it gives agents broad autonomous signup, email, OTP, browser, vault, and 2FA authority without enough user-control boundaries.

Install only if you want an agent to autonomously create and use Lumbox email infrastructure for signup and verification flows. Treat the LUMBOX_API_KEY, IMAP passwords, OTPs, magic links, and vault contents as sensitive credentials; confirm before allowing account creation, outbound email, browser automation, or credential storage, and prefer scoped inbox keys for subagents.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Findings (4)

Description-Behavior Mismatch

Medium
Confidence
90% confidence
Finding
The API reference explicitly expands the skill from simple email handling into browser automation, credential vault access, and TOTP/2FA tooling. In an agent context, this materially increases the privilege and attack surface of the integration, enabling account takeover workflows, secret access, and multi-step autonomous actions beyond what a user may expect from an email skill.

Vague Triggers

Medium
Confidence
92% confidence
Finding
The activation text is overly broad for a capability set that can autonomously self-sign up, provision inboxes, receive OTPs, and support account creation flows. In practice, this can cause an agent to invoke the skill in high-risk identity, signup, and communications contexts without an explicit user decision, increasing the chance of unauthorized account creation, verification handling, or abuse of third-party services.

Missing User Warnings

High
Confidence
98% confidence
Finding
The skill does not prominently warn that it will self-provision an external service account, persist an API key to a local file, and handle sensitive email and OTP data on the user's behalf. That omission is dangerous because users and orchestrating agents may not realize they are delegating identity-linked actions and secret management to a third-party service, which can lead to silent account creation, credential exposure, or unintended email actions.

Missing User Warnings

Medium
Confidence
84% confidence
Finding
The documentation describes returning IMAP credentials and plaintext API keys, including one-time key material, without strong handling guidance or safeguards. In an agent ecosystem, exposing secrets this way increases the chance they are logged, persisted in transcripts, forwarded to sub-agents, or otherwise mishandled, which can lead to inbox compromise and unauthorized API use.

VirusTotal

VirusTotal findings are pending for this skill version.

View on VirusTotal

Static analysis

No suspicious patterns detected.