baoyu-cover-image

Pass

Audited by Gen Agent Trust Hub on Apr 8, 2026

Risk Level: SAFEPROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
  • [PROMPT_INJECTION]: The skill contains instructions in 'references/base-prompt.md' that attempt to override standard safety filter refusals regarding copyrighted content, instructing the model to provide stylistic alternatives instead.
  • [PROMPT_INJECTION]: The skill has an indirect prompt injection surface as it ingests untrusted article data into generation prompts. Ingestion points: User-provided article files and reference image descriptions. Boundary markers: Uses headers such as '# Content Context' in 'references/workflow/prompt-template.md'. Capability inventory: Writing prompts and images to the file system and calling external generation skills. Sanitization: No sanitization of ingested content was observed.
  • [COMMAND_EXECUTION]: The skill uses shell commands in 'SKILL.md' to verify the presence of configuration files in both the project directory and the user's home directory.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 8, 2026, 06:31 AM