ClawHub
Back to skill

Security audit

Video Editing Agent (VEA)

Security checks across malware telemetry and agentic risk

Overview

VEA appears to be a legitimate video-editing skill, but it asks for powerful installation, credential, and third-party media-processing access that users should review carefully.

Install only if you are comfortable giving the service access to selected videos and the listed provider credentials. Use dedicated or least-privilege API keys, avoid broad personal Google application-default credentials when possible, inspect or pin the external repo and uv installer before running them, confirm each processing job before media leaves your machine, and delete local outputs/indexes after sensitive projects.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
  • Tool MisuseTool Parameter Abuse, Chaining Abuse, Unsafe Defaults
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • MCP Least PrivilegeUnderdeclared Capability, Wildcard Permission, Missing Permission Declaration
Findings (7)

Lp3

Medium
Category
MCP Least Privilege
Confidence
95% confidence
Finding
The skill clearly instructs the agent/user to run shell commands, install software, invoke Python, curl endpoints, and use ffmpeg, yet it declares no permissions to reflect those capabilities. This mismatch weakens policy enforcement and user understanding, making it easier for a seemingly harmless skill to trigger command execution paths without explicit authorization boundaries.

Vague Triggers

Medium
Confidence
86% confidence
Finding
The activation text is very broad ('use when asked to ... perform any video editing task'), which can cause the skill to be selected in many situations involving arbitrary media files, shell-driven processing, and external API usage. Over-broad routing increases the chance that sensitive files or unintended requests are handed to a powerful skill without narrow task scoping or additional consent.

Missing User Warnings

Medium
Confidence
99% confidence
Finding
The installation flow includes piping a remotely fetched script directly into the shell without any integrity verification or warning. If the upstream server, CDN, DNS, TLS termination, or install script is compromised, this results in immediate arbitrary code execution on the host.

Missing User Warnings

Medium
Confidence
92% confidence
Finding
The documentation explicitly describes sending video, transcripts, scripts, subtitles, and music-selection data to multiple third-party services, but it provides no privacy notice, consent guidance, data minimization expectations, or retention/sharing caveats. In a video-editing skill, user content is often highly sensitive, so omitting these warnings increases the risk of accidental unauthorized data disclosure and noncompliant handling of personal or proprietary media.

External Transmission

Medium
Category
Data Exfiltration
Content
```bash
# Download from Soundstripe API
SOUNDSTRIPE_KEY=$(jq -r '.api_keys.SOUNDSTRIPE_KEY' ~/vea/config.json)
curl -s "https://api.soundstripe.com/v1/songs/TRACK_ID" \
  -H "Authorization: Token $SOUNDSTRIPE_KEY" | jq '.included[0].attributes.versions.mp3'

# Mix with ffmpeg (15-20% music volume)
Confidence
84% confidence
Finding
https://api.soundstripe.com/

External Script Fetching

Low
Category
Supply Chain
Content
cd vea-open-source

# Install uv package manager
curl -LsSf https://astral.sh/uv/install.sh | sh

# Install dependencies
uv sync
Confidence
98% confidence
Finding
curl -LsSf https://astral.sh/uv/install.sh | sh

Chaining Abuse

High
Category
Tool Misuse
Content
cd vea-open-source

# Install uv package manager
curl -LsSf https://astral.sh/uv/install.sh | sh

# Install dependencies
uv sync
Confidence
99% confidence
Finding
| sh

VirusTotal

66/66 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.