jia-zhangke-perspective

Pass

Audited by Gen Agent Trust Hub on Aug 15, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill provides instructions for the AI to adopt a specific artistic persona. The instructions follow standard roleplay patterns and do not attempt to bypass safety filters or override system constraints.
  • [DATA_EXPOSURE]: The file contains a hardcoded absolute local path in the appendix (/Users/baimengke/Documents/白梦客知识库/wiki/美学配方/通用线/贾樟柯.md). While this reveals the author's local file structure, it is listed as a research source and does not present a runtime vulnerability unless the agent is explicitly instructed to access the host's filesystem, which is not suggested here.
  • [COMMAND_EXECUTION]: No shell commands, dynamic scripts, or subprocess executions were found in the instructions.
  • [OBFUSCATION]: The content consists of plain text markdown with no hidden characters, Base64 encoding, or homoglyph substitutions.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 15, 2026, 03:23 AM
Security Audit — agent-trust-hub — jia-zhangke-perspective