klingai-hello-world
Pass
Audited by Gen Agent Trust Hub on Sep 9, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes structured JSON responses from the Kling AI API (api.klingai.com).
- Ingestion points: API responses are parsed in the Python and Node.js examples within SKILL.md and references/examples.md.
- Boundary markers: The Instructions section in SKILL.md explicitly directs the agent to run requests in a sandbox and verify rights and policy outcomes.
- Capability inventory: The skill is granted Bash, Write, Read, and Edit tools.
- Sanitization: The code examples perform standard JSON parsing (e.g., jq or response.json()) to extract specific fields like job_id and video_url.
- [EXTERNAL_DOWNLOADS]: The skill provides examples (e.g., in references/examples.md) for downloading generated video files from the official Kling AI content delivery network (cdn.klingai.com).
Audit Metadata