baoyu-xhs-images
Pass
Audited by Gen Agent Trust Hub on Mar 16, 2026
Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill executes bash commands to verify the presence of configuration files in the local project and user home directories (SKILL.md).
- [COMMAND_EXECUTION]: It uses npx and the bun runtime to execute internal image generation scripts (references/workflows/prompt-assembly.md).
- [PROMPT_INJECTION]: Vulnerable to indirect prompt injection. 1. Ingestion points: Untrusted user-provided articles or text are analyzed in Step 1. 2. Boundary markers: Uses template placeholders such as {CONTENT_SECTION} within prompts. 3. Capability inventory: Shell command execution, script execution, and file system write access for saving analysis, prompts, and images. 4. Sanitization: No sanitization or filtering of the input content is documented or performed before processing.
Audit Metadata