ClawHub

cloudbypass

Security checks across malware telemetry and agentic risk

Overview

This skill openly routes protected-site requests through Cloudbypass, but its broad request forwarding and write-method helpers need careful review before use.

Install only for targets you are authorized to access through a third-party bypass provider. Use a dedicated Cloudbypass key and proxy, avoid forwarding account cookies or authorization headers unless explicitly approved, restrict destinations, monitor usage and billing, and require manual review before POST, PUT, DELETE, downloads, or autonomous browsing.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (3)

Description-Behavior Mismatch

Medium
Confidence
94% confidence
Finding
The skill exposes generic POST, PUT, and DELETE helpers even though the stated purpose is protected-page retrieval. In combination with a service explicitly designed to bypass Cloudflare/Turnstile protections, this enables arbitrary state-changing requests against protected targets through a third-party relay, materially expanding abuse potential beyond passive retrieval.

Missing User Warnings

Medium
Confidence
97% confidence
Finding
The request method forwards arbitrary target URLs, headers, bodies, and cookies to the external Cloudbypass service, effectively proxying sensitive session material and full request contents to a third party. In this skill's context, that is especially dangerous because the service is marketed for bypassing anti-bot protections, so the code facilitates covert access and exfiltration of authenticated or sensitive traffic without guardrails.

Natural-Language Policy Violations

High
Confidence
99% confidence
Finding
The comments explicitly state the skill is intended to bypass Cloudflare and similar anti-bot protections. That stated purpose is itself a major red flag: it frames the skill as an evasion tool, and the rest of the implementation operationalizes that intent by relaying requests to a specialized bypass API with support for challenge-solving and protected resource access.

VirusTotal

64/64 vendors flagged this skill as clean.

View on VirusTotal