skills/tw93/waza/write/Gen Agent Trust Hub

write

Pass

Audited by Gen Agent Trust Hub on May 16, 2026

Risk Level: SAFE
Full Analysis
  • [EXTERNAL_DOWNLOADS]: References official Chinese copywriting guidelines from a public GitHub repository to ensure proper formatting of mixed-language text.
  • [COMMAND_EXECUTION]: Utilizes the GitHub CLI to fetch previous release notes, ensuring that generated changelogs match the project's established style and density.
  • [DATA_EXFILTRATION]: Accesses local project documentation and user-defined style summaries to personalize writing output. This data is used solely for the rewriting process within the local context.
  • [PROMPT_INJECTION]: Features a 'Document Review Mode' that identifies and flags personally identifiable information (PII) to help users avoid accidental data exposure.
Audit Metadata
Risk Level
SAFE
Analyzed
May 16, 2026, 02:06 PM