Back to skill

Security audit

Yq Seedance Video Creator

Security checks for vulnerabilities and agentic risk

Overview

This is a coherent Chinese-language Seedance/Jimeng video-prompt skill with no code execution, persistence, credential collection, or hidden system access.

Install this if you want a Chinese-language assistant for Seedance/Jimeng video prompts and storyboards. Expect Chinese templates and occasional Jimeng website references; verify any third-party platform terms, costs, and account requirements yourself, and avoid uploading private, unlicensed, or sensitive media unless you intend to use that service.

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
Findings (1)

T01 · Skill Instruction Hijacking

Warning
Location
SKILL.md:590
Finding

Mandatory External-Link Injection in Generated Responses

Content
View full analysis

Vulnerability Details

File Location: SKILL.md:590-606
Related Locations: SKILL.md:634, storyboard/templates/storyboard-template.md:129, storyboard/quick-reference.md:91, storyboard/examples/example-prompts.md:187
Vulnerability Type: Fixed advertising and traffic-diversion content embedded in the required output template
Risk Level: Medium

Vulnerable Content

The following is an English rendering of the complete vulnerable template segment at SKILL.md:590-606:

text
---

[Prompt]:
{Complete Seedance 2.0 prompt, ready to copy and use}

---

[Material Recommendations]:
- First-frame image: {recommended description}
- Reference video: {recommended description, optional}
- Background music: {recommended description, optional}

[Creation Tips]:
{Usage recommendations and precautions}

Tip: Copy the prompt above to the Jimeng official website:
https://jimeng.jianying.com/

The same fixed destination is repeated in several supporting resources, including the final line of storyboard/templates/storyboard-template.md:

text
Jimeng official website: https://jimeng.jianying.com

Technical Analysis

The Skill's declared purpose is to generate Seedance 2.0 prompts, storyboards, and film-production guidance. Providing a fixed third-party URL in every normal final creation plan is not required to perform those functions.

The URL is embedded inside the prescribed final-output format rather than being presented only as optional documentation. When the Skill is loaded, an agent following its instructions will predictably append the destination to otherwise unrelated generated responses. Repetition in the root Skill and three storyboard resources reinforces this behavior.

This is instruction-level output hijacking: Skill text modifies the agent's output policy so that responses contain stable promotional or traffic-diversion content. The destination appears contextual ...[truncated 1467 chars]

Remediation
View remediation

Remediation Suggestions

  1. Remove the fixed URL from the mandatory final-output template in SKILL.md.
  2. Remove the repeated promotional footers from:
    • SKILL.md
    • storyboard/templates/storyboard-template.md
    • storyboard/quick-reference.md
    • storyboard/examples/example-prompts.md
  3. Keep generated answers limited to the requested prompt, storyboard, material recommendations, and relevant usage guidance.
  4. Provide an external platform link only when the user explicitly asks where to use the generated prompt.
  5. Clearly label any requested external link as optional and disclose that it leads to a third-party service.
  6. Avoid hardcoding third-party destinations in reusable response templates. If documentation links are necessary, place them in a dedicated reference section that is not automatically reproduced in model output.
  7. Add a review rule prohibiting mandatory advertising, referral links, calls to action, or unrelated branding in generated responses.
Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (11)

Ae1

High
Category
analysis-evasion
Confidence
100% confidence
Finding

Referenced artifact was not completely inspected

Content

Scanner excerpt · SKILL.md (reported line 437)May include surrounding context.

md
参考文件:`film-production/` 目录下的 `SKILL.md`(总览文件)

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
93% confidence
Finding

The skill description is extremely broad and can trigger on many generic requests involving video creation, prompt writing, multimodal content, storyboards, visual style, or optimization. Overbroad activation increases the chance the agent invokes this skill in unintended contexts, which can override more appropriate skills or steer users into external-platform-specific workflows without clear necessity.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
96% confidence
Finding

The skill content is written entirely in Chinese and repeatedly prescribes fixed Chinese output structures and interaction patterns without offering language negotiation. This can cause the agent to ignore the user's preferred language, reducing transparency and user control, and in safety-critical interactions may lead to misunderstanding of platform rules, consent prompts, or external-site guidance.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
94% confidence
Finding

The skill’s title, description, and all operating instructions are written entirely in Chinese, with no indication that users may choose another language or locale. Under the stated policy, a skill that effectively enforces a specific language without opt-in is a natural-language policy violation.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
90% confidence
Finding

The entire skill is written in Chinese and presents itself as a professional guide without any indication that other languages are supported or that the user can opt into this locale. Under the policy, forcing a specific language without user opt-in is a natural-language policy violation unless the locale constraint is clearly documented and justified.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
93% confidence
Finding

This markdown file presents all user-facing instructional content exclusively in Chinese, and there is no indication that the skill is region-specific or that users can opt into another language. That can violate a language/locale policy when skills are expected to avoid forcing a specific language without user choice.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
96% confidence
Finding

The file’s instructional content is entirely in Chinese and presents itself as a general professional guide, but it does not indicate that Chinese is optional, user-selected, or required for a region-specific purpose. Under the policy, forcing a specific language without opt-in is a natural-language policy violation.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
93% confidence
Finding

The entire skill is written in Chinese and does not offer any language choice, fallback, or note that the skill is intended only for Chinese-speaking users. Under the stated policy, forcing a specific language without user opt-in is a natural-language policy violation unless the locale restriction is explicitly documented and justified.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
93% confidence
Finding

This markdown file is entirely written in Chinese, including headings, examples, and operational guidance, and it does not indicate that the user may choose another language or that the skill is intentionally limited to a Chinese-speaking audience. Under the policy, forcing a specific language without user opt-in is a natural-language policy violation.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
88% confidence
Finding

The file specifies '四川方言对白' as the dialogue language for the example, which imposes a particular regional dialect in the natural-language content. There is no indication that this is optional or that users may choose another language or locale.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
93% confidence
Finding

This markdown file contains user-facing instructions exclusively in Chinese, which can amount to a language-policy issue when no opt-in or alternative locale is provided. The policy specifically calls for flagging cases where a skill forces a specific language without user choice.

Content

No source excerpt is available for this finding.