writer-Pro-one
Pass
Audited by Gen Agent Trust Hub on Jun 1, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill provides a comprehensive framework for stylistic text transformation and translation. All instructions are focused on creative writing and prose optimization, with no evidence of malicious intent or unauthorized actions.
- [COMMAND_EXECUTION]: The instructions do not direct the agent to execute shell commands or interact with the operating system. A reference to Windows PowerShell's 'Get-Content' is provided as a courtesy to human users for correct file encoding and is not an instruction for the agent to execute.
- [DATA_EXFILTRATION]: There are no network operations, external URL fetches, or attempts to access sensitive files (such as .ssh, .aws, or .env). The skill functions entirely through text processing within the agent's provided context.
- [PROMPT_INJECTION]: Instructions regarding the removal of 'AI writing traces' and avoiding specific templates are stylistic constraints intended to improve output quality. These do not constitute an attempt to bypass the model's safety guardrails or ethical protocols.
Audit Metadata