Back to skill

Security audit

Ai Short Drama Director

Security checks for vulnerabilities and agentic risk

Overview

The skill is a coherent AI short-drama production workflow, but it includes an unscoped system package install command that users should review before use.

Install only if you are comfortable with a workflow that generates many local media files and may ask the agent to install ffmpeg at the system level. Prefer using it in a disposable or project-specific workspace, and require explicit approval before any apt-get command is run.

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
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • 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)

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
89% confidence
Finding

The manifest uses very broad trigger phrases such as 'drama', 'storyboard', and '短剧/剧本', which can cause the skill to activate for many ordinary creative requests. Over-broad activation is dangerous because this skill performs extensive file creation and multi-step media generation, so unintended invocation can lead to unnecessary tool use, workspace modification, and user confusion.

Content

No source excerpt is available for this finding.

Context-Inappropriate Capability

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The skill explicitly instructs running apt-get update && apt-get install -y ffmpeg, which introduces system-level package installation during normal operation. Even though ffmpeg is relevant to video production, automatically directing an agent to modify the host environment increases attack surface, can fail unpredictably across environments, and may violate sandbox or least-privilege expectations.

Content

No source excerpt is available for this finding.

Missing User Warnings

Low
Category
Not specified by scanner
Confidence
91% confidence
Finding

The skill description and workflow do not prominently warn users that it will create many directories and write numerous intermediate and final files under output/. In context, the skill performs broad workspace mutations across six stages, so lack of upfront disclosure can lead to surprising side effects and accidental execution in sensitive or cluttered working directories.

Content

No source excerpt is available for this finding.

Static analysis

No suspicious patterns detected.