huashu-script-polish

Pass

Audited by Gen Agent Trust Hub on Feb 21, 2026

Risk Level: SAFENO_CODE
Full Analysis
  • [NO_CODE] (SAFE): The skill consists solely of markdown-formatted instructions and lacks any executable scripts, binaries, or configuration files that initiate code execution.
  • [REMOTE_CODE_EXECUTION] (SAFE): No patterns for downloading or executing remote code were found.
  • [DATA_EXFILTRATION] (SAFE): No network-related commands or potential data-leakage vectors are present.
  • [PROMPT_INJECTION] (SAFE): The instructions do not contain commands to bypass safety filters or ignore system rules.
  • [DATA_EXPOSURE] (SAFE): The skill references local informational files but does not target sensitive system paths or credentials.
  • [INDIRECT_PROMPT_INJECTION] (LOW): The skill processes user-supplied scripts for editing. 1. Ingestion points: User-provided script text. 2. Boundary markers: Absent. 3. Capability inventory: None (the skill has no code or network access). 4. Sanitization: Absent.
Audit Metadata
Risk Level
SAFE
Analyzed
Feb 21, 2026, 07:26 AM