klingai-async-workflows

Pass

Audited by Gen Agent Trust Hub on Apr 4, 2026

Risk Level: SAFEPROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
  • [PROMPT_INJECTION]: The skill incorporates user-provided text prompts directly into API requests to Kling AI without sanitization or the use of boundary markers (found in SKILL.md and references/workflow-implementation.md). This creates a surface for indirect prompt injection where untrusted input could attempt to manipulate model behavior.\n
  • Ingestion points: Prompt data enters through function arguments in SKILL.md and job objects in references/workflow-implementation.md.\n
  • Boundary markers: The skill does not use delimiters or instructions to isolate user prompts from the rest of the request structure.\n
  • Capability inventory: The skill can perform network operations (via requests and aiohttp) and write to the local file system (seen in SKILL.md).\n
  • Sanitization: No evidence of input filtering or escaping was found in the provided scripts.\n- [EXTERNAL_DOWNLOADS]: The skill contains logic to download video files from URLs generated by the Kling AI API and save them to the local disk (found in SKILL.md). This is a standard and expected operation for the skill's intended purpose.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 4, 2026, 05:48 PM