klingai-install-auth
Pass
Audited by Gen Agent Trust Hub on Sep 9, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill provides code templates that ingest user-supplied text to interact with the Kling AI video generation API. This constitutes a potential surface for indirect prompt injection if an agent passes untrusted external data into these functions without prior validation.
- Ingestion points: Found in
references/environment-setup.mdandSKILL.mdwithin thecreate_video(prompt)andtext2videocode examples. - Boundary markers: No explicit delimiters or boundary markers are included in the code snippets to separate user prompts from system instructions.
- Capability inventory: The functions utilize the
requests,axios, andhttpxlibraries to perform network POST operations tohttps://api.klingai.com/v1. - Sanitization: The example code does not implement sanitization or filtering of the
promptvariable before interpolation into the JSON payload.
Audit Metadata