grammar-style-enhancer
Pass
Audited by Gen Agent Trust Hub on Feb 16, 2026
Risk Level: LOWNO_CODE
Full Analysis
- [Prompt Injection] (SAFE): No instructions identified that attempt to override AI safety constraints, bypass instructions, or extract system prompts.
- [Data Exposure & Exfiltration] (SAFE): No references to sensitive file paths, environment variables, or network-bound operations were detected.
- [Indirect Prompt Injection] (INFO): The skill is designed to process untrusted text input for editing. However, it lacks dangerous capabilities such as file writing, shell execution, or network requests, which limits the potential impact of adversarial content to the generated text response only.
- [Remote Code Execution] (SAFE): The skill does not reference any external scripts, package managers, or dynamic execution functions.
Audit Metadata