Back to skill

Security audit

muse-video

Security checks for vulnerabilities and agentic risk

Overview

The skill provides a coherent video-generation workflow with disclosed media uploads, controlled authentication, and limits on paid retries.

Your prompts and selected images will be sent to RunAPI, and generation may incur charges. Review its privacy terms before using sensitive media and protect your API key. External dependency behavior was not audited, and the linked live model documentation could not be retrieved.

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 workflow uploads local media to an external service without requiring a user-facing privacy warning.

Medium
Category
Output Handling
Confidence
85% confidence
Finding

Line 51 directs the agent to supply local media paths for automatic upload or to perform a separate upload, and line 64 submits the request. Although the upload mechanism is described to the agent, the skill never instructs it to warn the user that their source media and prompt will be transmitted to RunAPI for processing.

Content

No source excerpt is available for this finding.