image-generation

Pass

Audited by Gen Agent Trust Hub on Mar 11, 2026

Risk Level: SAFE
Full Analysis
  • [DATA_EXPOSURE]: The skill adheres to security best practices by utilizing environment variables (DOUBAO_API_KEY) and configuration templates to avoid hardcoding sensitive credentials. The local 'config.yaml' is correctly excluded via documentation and templates.
  • [EXTERNAL_DOWNLOADS]: The skill communicates exclusively with official Volcengine (ByteDance) API endpoints at 'ark.cn-beijing.volces.com' for image generation and downloads, which is a recognized and well-known service provider.
  • [INDIRECT_PROMPT_INJECTION]: The skill provides a surface for indirect prompt injection as it processes user-provided descriptions and file paths.
  • Ingestion points: CLI arguments '--prompt' and '--output' in 'scripts/doubao_image_gen.py'.
  • Boundary markers: Absent.
  • Capability inventory: The script has the ability to fetch data from the network ('requests.get') and write to the local filesystem ('open().write()').
  • Sanitization: Input file paths are used directly for asset storage, which is consistent with the skill's primary purpose of image asset management.
  • [COMMAND_EXECUTION]: The skill relies on standard Python script execution to interact with the Doubao API, with no evidence of arbitrary or malicious command execution.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 11, 2026, 06:52 AM