humanizer

Pass

Audited by Gen Agent Trust Hub on May 5, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADS
Full Analysis
  • [SAFE]: The skill implements a set of writing guidelines to reduce AI-specific markers in text. It uses standard tools to process content provided by the user.
  • [EXTERNAL_DOWNLOADS]: The skill references a Wikipedia page for user reference. Wikipedia is a well-known and trusted information service.
  • [PROMPT_INJECTION]: The skill processes user-supplied text for rewriting, creating an attack surface for indirect prompt injection. 1. Ingestion points: User input or files specified by the user in SKILL.md. 2. Boundary markers: Absent. 3. Capability inventory: Read, Write, Edit, Grep, Glob, and AskUserQuestion across all functions. 4. Sanitization: Absent. While this is an attack surface, the skill instructions are focused on linguistic humanization, making it a low-risk utility.
Audit Metadata
Risk Level
SAFE
Analyzed
May 5, 2026, 10:43 AM
Security Audit — agent-trust-hub — humanizer