seedream-5-pro-prompt

Pass

Audited by Gen Agent Trust Hub on Aug 26, 2026

Risk Level: SAFECOMMAND_EXECUTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill provides a Python script scripts/validate_prompt.py and instructs the agent to execute it using python scripts/validate_prompt.py <prompt-package.json> to verify the integrity of prompt packages. The script is contained within the skill package and uses standard library modules for structural validation.
  • [INDIRECT_PROMPT_INJECTION]: The skill is designed to analyze and extract text and annotations from user-supplied images as part of its core workflow (references/image-analysis.md), which represents an indirect prompt injection surface.
  • Ingestion points: User-provided images and screenshots processed according to references/image-analysis.md.
  • Boundary markers: Absent; there are no explicit delimiters or instructions to ignore malicious text embedded within the analyzed images.
  • Capability inventory: Shell command execution via the bundled validation script.
  • Sanitization: Absent; the skill is instructed to record text found in images verbatim.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 26, 2026, 10:28 AM
Security Audit — agent-trust-hub — seedream-5-pro-prompt