Back to skill

Security audit

短剧视频提示词

Security checks for vulnerabilities and agentic risk

Overview

ClawScan rates this skill benign with medium confidence: the completed AIG scan reported no findings, while Skillspector flagged language restrictions and a missing privacy warning for external media processing; its coverage was partial.

The verdict is benign, not a guarantee of safety. Before installing, consider whether fixed language requirements and the missing privacy warning fit your needs. Skillspector inspected all 12 components, but its analysis was partial because two local references could not be resolved; the evidence is therefore incomplete.

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 (5)

The skill mandates English visual descriptions without user opt-in or an explicit technical justification.

Medium
Category
Output Handling
Confidence
91% confidence
Finding

The instruction requires English for both section labels and visual descriptions. Fixed API labels may be necessary, but the file does not explain why free-form descriptions must be English or offer a language choice.

Content

No source excerpt is available for this finding.

Reference-media submission instructions omit a privacy warning about third-party processing.

Medium
Category
Output Handling
Confidence
90% confidence
Finding

The workflow instructs the agent to submit character voice references and visual reference images to the generation interface, rather than merely listing supported tools. It does not warn that potentially identifying voice recordings and images will be processed by an external service.

Content

No source excerpt is available for this finding.

Dialogue instructions require Chinese without offering a user language choice.

Medium
Category
Output Handling
Confidence
96% confidence
Finding

The instruction specifies that dialogue must use the form “角色用中文说” and maintain a single spoken language, imposing Chinese without user opt-in. No model limitation or regional requirement justifies this restriction; line 32 also provides guidance for English prompts.

Content

No source excerpt is available for this finding.

The dialogue template requires Chinese speech without user opt-in.

Medium
Category
Output Handling
Confidence
97% confidence
Finding

The prescribed dialogue format, {角色用中文说:“逐字台词”}, explicitly instructs characters to speak Chinese. No language choice or technical justification for this restriction is provided; the recommended zh-CN configuration on line 9 reinforces the language assumption.

Content

No source excerpt is available for this finding.

The recommended configuration specifies Simplified Chinese for generated video prompts without user opt-in.

Low
Category
Output Handling
Confidence
88% confidence
Finding

The video_prompt_language: "zh-CN" setting explicitly selects a natural language for generated prompts. The file does not offer a language choice or explain why Wan 3.0 technically requires this locale; selecting the model alone does not establish a user language preference.

Content

No source excerpt is available for this finding.