The Agent Skills Directory

[SAFE]: Document Automation API Usage. The skill utilizes standard office automation interfaces (Application.Selection) to read and modify text currently selected by the user. These operations are scoped to the host application's document context and do not interact with the underlying operating system or external networks.
[SAFE]: No Malicious Patterns Detected. A comprehensive scan for prompt injection, obfuscation, hardcoded credentials, and persistence mechanisms yielded no findings. The instructions strictly adhere to the stated purpose of text refinement.
[SAFE]: Processing of Untrusted Input. While the skill processes user-selected text, its capabilities are restricted to document selection replacement. There is no evidence of network exfiltration or dangerous command execution, mitigating the risk of processing untrusted content.

rewrite-polish