klingai-common-errors

Pass

Audited by Gen Agent Trust Hub on Sep 9, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill functions as a troubleshooting guide and reference for Kling AI API status codes and task failures, containing no malicious patterns.
  • [DATA_EXPOSURE_AND_EXFILTRATION]: The instructions include defensive measures against data exposure, specifically requiring the redaction of authorization headers, prompts, and source assets before generating diagnostic receipts.
  • [INDIRECT_PROMPT_INJECTION]: The skill handles untrusted data from API responses but mitigates risks through mandatory evidence redaction and strict boundary enforcement.
  • Ingestion points: API response bodies and task status messages in SKILL.md.
  • Boundary markers: Explicit instructions to exclude prompts and assets from final output.
  • Capability inventory: Bash, Read, Write, Edit, and Grep tools.
  • Sanitization: Mandatory redaction rules and cleanup procedures are defined to handle sensitive material in diagnostics.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 9, 2026, 03:46 AM
Security Audit — agent-trust-hub — klingai-common-errors