klingai-debug-bundle
Pass
Audited by Gen Agent Trust Hub on Sep 9, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes data returned from the Kling AI API (
api.klingai.com). As with any tool that ingests external, potentially untrusted content, there is a theoretical surface for indirect prompt injection where malicious API payloads could attempt to influence the agent's logic. This risk is managed by the skill's focus on structured logging and the use of synthetic fixtures. - [SAFE]: The skill demonstrates proper credential management by accessing API keys via environment variables (
KLING_ACCESS_KEY,KLING_SECRET_KEY) rather than including hardcoded secrets. - [SAFE]: Documentation specifically mandates the use of synthetic data and the application of redaction policies to logs, mitigating the risk of accidental sensitive data exposure during troubleshooting sessions.
Audit Metadata