Back to skill

Security audit

ai-agentic-video-editor

Security checks for vulnerabilities and agentic risk

Overview

The skill is marked suspicious because its cloud editing workflow sends briefs and media to Levea and calls for publicly accessible media URLs without a privacy warning; no malicious behavior was reported, but scan coverage was partial.

Before installing, consider whether you are comfortable sending briefs and media to Levea and making media URLs publicly accessible. Avoid using it with sensitive material unless you understand and authorize that sharing; the scan was partial, so it does not establish that all referenced content was reviewed.

Vulnerability Patterns
  • Skill Instruction HijackingAlters the agent's session goals or safety constraints when the skill loads
  • Agent Memory PoisoningWrites attacker-controlled rules into memory that affect later sessions
  • Remote Payload Retrieval and ExecutionFetches external code whose behavior can change after review
  • Embedded Malicious CodeShips malicious scripts inside the skill and executes them locally
  • Unauthorized Access and Privilege EscalationObtains permissions beyond the task's legitimate needs
Vulnerability Patterns
  • Output HandlingUnvalidated Output Injection, Cross-Context Output, Unbounded Output
  • 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 (1)

The cloud-editing workflow lacks a privacy warning for sending user briefs and media to Levea and requiring publicly accessible media URLs.

Medium
Category
Output Handling
Confidence
89% confidence
Finding

The skill directs agents to send user briefs and media URLs to an external editing service and specifies publicly accessible input URLs, without requiring a sensitivity check or informed authorization for that disclosure. This could expose confidential footage, personal information, or proprietary material to the service and to anyone who obtains an accessible media URL. The explicit cloud-service setup makes this an inadequately guarded, intended data transfer rather than evidence of covert exfiltration.

Content

No source excerpt is available for this finding.