kling-ai
Pass
Audited by Gen Agent Trust Hub on Nov 11, 2026
Risk Level: SAFENO_CODEEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
- [NO_CODE]: The skill is entirely informational, consisting of markdown guides, pricing reference tables, and prompt templates. It does not contain any executable scripts, binaries, or logic that performs operations on the host system.
- [EXTERNAL_DOWNLOADS]: The documentation references the installation of the official Kling AI CLI tool (
@klingai/cli-global) from the NPM registry. As Kling AI is a well-known AI service, this reference to its official utility is considered safe and standard practice for developer-oriented skills. - [INDIRECT_PROMPT_INJECTION]: The skill defines workflows for generating content based on user-provided prompts and external media assets (images, videos, and voices). This establishes a data ingestion surface inherent to AI generation tools.
- Ingestion points: User-supplied text prompts and media references (@ImageN, @VideoN, etc.) processed by the Kling model.
- Boundary markers: The skill encourages the use of structured prompt blocks (e.g., [Asset Roles], [Sequence of Events]) to maintain control over how the AI interprets different inputs.
- Capability inventory: The skill facilitates interaction with the external Kling AI API and CLI for video generation tasks.
- Sanitization: The skill assumes the use of Kling AI's internal safety filters and does not specify local sanitization logic.
Audit Metadata