Skills
Navigate
skills/bytedance/agentkit-samples/byted-byteplus-vod-video-enhancement/Gen Agent Trust Hub

byted-byteplus-vod-video-enhancement

Pass

Audited by Gen Agent Trust Hub on Sep 21, 2026

Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill executes Python scripts including upload.py, quality_enhance.py, and poll_execution.py to interact with the BytePlus VOD API.
  • [INDIRECT_PROMPT_INJECTION]: The skill processes external data such as public URLs for pull uploads and JSON-formatted parameters for video enhancement configurations.
  • Ingestion points: scripts/upload.py (URL input), scripts/quality_enhance.py (JSON arguments/parameter files).
  • Boundary markers: Absent.
  • Capability inventory: Restricted local file reading, network POST/GET operations to BytePlus API endpoints (vod.byteplusapi.com).
  • Sanitization: The _validate_local_path function in scripts/upload.py restricts file access to the workspace, /tmp, and explicit allowed directories; JSON inputs are parsed and validated against white-listed configuration values.
  • [SAFE]: The implementation of _validate_local_path is a significant security control that prevents the unauthorized upload of sensitive files outside the agent's defined scope. Additionally, the skill utilizes environment variables for secrets as per platform best practices.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 21, 2026, 03:15 PM