baoyu-article-illustrator
Pass
Audited by Gen Agent Trust Hub on May 25, 2026
Risk Level: SAFE
Full Analysis
- [PROMPT_INJECTION]: The skill uses instructional guidance to manage its multi-step illustration workflow. While it directs the agent to generate stylistic alternatives for sensitive subjects in image prompts to ensure process continuity, this is a narrow instruction that does not compromise overall agent safety filters.- [COMMAND_EXECUTION]: Shell commands are used strictly for environment checks, such as identifying the location of preference files (
test -f). These are standard, safe operations for configuration discovery.- [DATA_EXFILTRATION]: The skill interacts with local project files and user settings at documented paths. It does not perform any unauthorized network communications or access sensitive system credentials like SSH keys or API tokens.- [SAFE]: The skill demonstrates secure practices by requiring explicit user confirmation before image generation and providing a detailed outline of planned illustrations for review.
Audit Metadata