ClawHub

Ai Video Workflow

Security checks across malware telemetry and agentic risk

Overview

This skill is mostly a coherent AI video workflow, but two bundled media helpers can fetch any URL and write it to any chosen local path, which is broader than the stated asset-download purpose.

Install only if you are comfortable with the skill running local Bun scripts that access provider API keys and write media files. Keep API keys in local config or environment variables, review download commands before execution, and direct outputs to a dedicated project folder rather than broad or sensitive filesystem paths.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • 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 (2)

Context-Inappropriate Capability

Medium
Confidence
93% confidence
Finding
The download command accepts any user-supplied URL and writes the response to any user-supplied output path without restricting the host, scheme, or file destination. In an agent skill context, this expands the tool from a Pexels helper into a generic fetch-and-write primitive, which can enable unintended external access, retrieval of untrusted content, or overwriting local files if upstream prompts or inputs are influenced by an attacker.

Missing User Warnings

Medium
Confidence
83% confidence
Finding
The code fetches any user-supplied URL and writes the response directly to any user-supplied local path, with no scheme/host restrictions, path safety checks, or confirmation step. In agent or automation contexts, this can enable SSRF-style internal network access, unexpected local file overwrite, or dropping attacker-controlled content onto disk.

VirusTotal

65/65 vendors flagged this skill as clean.

View on VirusTotal