web-asset-generator
Pass
Audited by Gen Agent Trust Hub on Feb 15, 2026
Risk Level: LOWEXTERNAL_DOWNLOADSPROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- EXTERNAL_DOWNLOADS (LOW): The skill requires the installation of third-party Python packages 'pilmoji' and 'emoji' (<2.0.0) from PyPI. These libraries are not from trusted organizations defined in the security policy, though they are standard for the intended functionality.
- PROMPT_INJECTION (LOW): Indirect Prompt Injection vulnerability (Category 8): Untrusted data enters the skill via the '--suggest' (project description) and '--text' (image overlay) arguments in the generator scripts. There is no evidence of boundary markers or sanitization in the provided metadata. However, because the skill's capabilities are limited to generating local image files and HTML tags, the risk to the agent's reasoning or external security is considered low.
- COMMAND_EXECUTION (LOW): The skill's operational model involves the agent executing local Python scripts to generate assets. While this is expected behavior for an asset generation skill, it represents a privilege that requires monitoring.
Audit Metadata