ClawHub

Agent Onboarding

Security checks across malware telemetry and agentic risk

Overview

This appears to be legitimate Bright Data onboarding, but it asks users to run a remote installer and persist API credentials with limited safety guidance.

Review before installing. Prefer the npm or npx path over piping a remote script into bash, or download and inspect/verify the installer first. Use a scoped Bright Data API key, protect local config and `.env` files, avoid exposing tokens in URLs/logs/screenshots, and review any additional skills before using `bdata skill add`.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
Findings (2)

Missing User Warnings

Medium
Confidence
93% confidence
Finding
The skill recommends a `curl ... | bash` installer as the fastest path without any warning or integrity verification guidance. Even if the vendor is legitimate, piping a remote script directly into a shell gives that remote endpoint immediate code execution on the host, which is a meaningful supply-chain and operator-safety risk.

Missing User Warnings

Medium
Confidence
84% confidence
Finding
The skill states that `bdata login` saves the API key locally and that users never need to paste a token again, but it does not explain where the credential is stored, what file permissions are expected, or how to protect or revoke it. In an agent-oriented setup, local secret persistence increases the chance of token exposure through logs, shared workspaces, backups, or overly broad filesystem access.

VirusTotal

65/65 vendors flagged this skill as clean.

View on VirusTotal