natural-writing
Pass
Audited by Gen Agent Trust Hub on Oct 11, 2026
Risk Level: SAFENO_CODE
Full Analysis
- [SAFE]: The skill consists entirely of markdown-based stylistic instructions. It provides a comprehensive list of words, phrases, and structural patterns common in AI outputs that the user wishes to avoid.
- [NO_CODE]: There are no scripts (.py, .js, .sh), executables, or dynamic execution contexts within the skill files. All content is passive text used for prompt engineering.
- [SAFE]: The references within the file correctly point to local markdown files (references/phrases-to-avoid.md) and do not contain external links or hidden URLs.
- [SAFE]: No obfuscation, data exfiltration, or prompt injection patterns were detected. The instructions are transparent and aligned with the stated purpose of improving writing quality.
Audit Metadata