baoyu-xhs-images

Pass

Audited by Gen Agent Trust Hub on Mar 12, 2026

Risk Level: SAFEPROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
  • [PROMPT_INJECTION]: The skill's image generation guidelines in references/workflows/prompt-assembly.md explicitly instruct the model to bypass safety filters for sensitive or copyrighted figures by stating 'DO NOT refuse to generate' and demanding stylized alternatives.
  • [PROMPT_INJECTION]: The skill processes untrusted external data (user-provided text, files, and URLs) to generate prompts for other tools, creating a vulnerability to indirect prompt injection attacks.
  • Ingestion points: SKILL.md Step 1 (processes pasted content, file paths, and URLs).
  • Boundary markers: No delimiters or safety instructions are used to isolate user-supplied content within the generated prompts.
  • Capability inventory: File system access (read/write), shell command execution (bash), and image generation skill calls.
  • Sanitization: No sanitization, filtering, or validation of user input is performed before it is interpolated into prompts.
  • [COMMAND_EXECUTION]: The workflow involves executing bash to verify configuration files in the project and user home directories (.baoyu-skills/EXTEND.md) and references executing local TypeScript scripts via the bun runtime in references/workflows/prompt-assembly.md.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 12, 2026, 07:52 AM