zhang-yiming-perspective

Pass

Audited by Gen Agent Trust Hub on Apr 9, 2026

Risk Level: SAFEPROMPT_INJECTIONCOMMAND_EXECUTIONEXTERNAL_DOWNLOADS
Full Analysis
  • [PROMPT_INJECTION]: The core instructions in SKILL.md direct the agent to suppress repetitive safety disclaimers and character-breaking refusals to maintain persona consistency, overriding default AI behavior patterns for the sake of roleplay.
  • [EXTERNAL_DOWNLOADS]: The skill incorporates an automated research workflow that ingests data from external web searches to generate responses, creating a surface for processing potentially malicious third-party content without specified sanitization.
  • [COMMAND_EXECUTION]: The README documentation provides shell commands for installation via npx, which involve executing remote scripts associated with the author's project repository.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 9, 2026, 07:12 AM