The Agent Skills Directory

[SAFE]: No malicious patterns or security vulnerabilities were identified during the analysis. The skill is limited to text manipulation within a specific local project structure.
[DATA_EXPOSURE_AND_EXFILTRATION]: File access is restricted to the local project directories for reading context and writing polished content. There are no network operations, hardcoded credentials, or access to sensitive system paths.
[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted input from local project files. However, the lack of dangerous capabilities like shell execution or web requests limits the risk, and the instructions explicitly enforce strict citation and scope constraints.

subsection-polisher