The Agent Skills Directory

[PROMPT_INJECTION]: No evidence of malicious prompt injection, jailbreaks, or safety filter bypasses was found. The skill contains strong instructions to the agent to ensure it is triggered during the writing process, but these are functional constraints for its intended purpose (text polishing) rather than attempts to override the system's core safety instructions.
[REMOTE_CODE_EXECUTION]: No remote code execution patterns or package manager installations (npm, pip, etc.) were detected. The skill specifically states it has no extra dependencies.
[DATA_EXFILTRATION]: No network operations (curl, wget, etc.) or sensitive file system access patterns were identified. The only external URL present is a link to the author's own GitHub repository for documentation and reference.
[COMMAND_EXECUTION]: No shell command execution or dynamic context injection (!command) was found. The slash commands shown in the documentation are examples for user interaction, not executable shell scripts.
[OBFUSCATION]: No obfuscated content, Base64 strings, zero-width characters, or hidden URLs were detected in any of the files.

de-ai-polish