baoyu-cover-image

Pass

Audited by Gen Agent Trust Hub on Jun 28, 2026

Risk Level: SAFEPROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
  • [PROMPT_INJECTION]: The skill contains instructions in 'references/base-prompt.md' that attempt to influence the behavior of image generation models by directing them not to refuse requests involving sensitive or copyrighted figures.\n- [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection because it extracts metadata (titles, summaries, and keywords) from user-provided articles and incorporates them directly into image generation prompts without sanitization or boundary markers.\n
  • Ingestion points: User-provided article content or files processed in SKILL.md Step 1.\n
  • Boundary markers: No explicit delimiters or instruction-blocking markers are used during prompt construction.\n
  • Capability inventory: Shell command execution for configuration checks and external tool calls for image generation.\n
  • Sanitization: No content validation or filtering is performed on input data.\n- [COMMAND_EXECUTION]: The skill uses basic shell commands like 'test' and 'echo' to verify the existence of its own configuration files in the project or home directory.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 28, 2026, 03:19 AM
Security Audit — agent-trust-hub — baoyu-cover-image