ClawHub
Back to skill

Security audit

Boxed FFmpeg

Security checks across malware telemetry and agentic risk

Overview

This skill appears intended for local media conversion, but it asks the agent to download and persistently run an unpinned remote WASM executable without integrity verification.

Review before installing. Use this only if you trust the openclaw-wasm-sandbox plugin and the GitHub-hosted WASM binary, or can independently verify a pinned release and checksum. Treat the downloaded WASM as executable code, and run it only on explicit media files you intend to process.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • 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
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Findings (3)

Intent-Code Divergence

Medium
Confidence
91% confidence
Finding
The skill claims processing is fully local and needs no network access, yet it instructs users to download an executable WASM component from a remote GitHub URL. This creates a supply-chain trust gap: the fetched binary could change, be replaced, or be tampered with, and the documentation downplays that risk by framing the workflow as local-only.

Vague Triggers

Medium
Confidence
86% confidence
Finding
Using the bare trigger phrase "ffmpeg" is overly broad and may activate the skill when a user is merely discussing FFmpeg rather than requesting this skill. Unintended activation can cause the agent to select this skill in the wrong context, increasing the chance of inappropriate file operations or surprising execution behavior.

Vague Triggers

Medium
Confidence
84% confidence
Finding
Generic triggers like "convert video" and "extract audio" are common user phrases and lack constraints tying them to this specific tool or to safe execution prerequisites. In an agent environment, such ambiguity increases the risk of accidental skill routing and unintended processing of workspace files.

VirusTotal

65/65 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.