The Agent Skills Directory

[SAFE]: The skill follows a highly structured and declarative workflow for processing academic reviewer comments. It includes explicit 'red lines' and core rules that strictly prevent the agent from fabricating data, results, statistical values, or citations.
[SAFE]: No remote code execution patterns, external package installations, or unauthorized network operations were found. The URLs referenced in the documentation are restricted to official Nature Portfolio and Springer Nature domains for providing editorial guidance.
[SAFE]: The skill demonstrates no signs of obfuscation, credential exposure, or persistence mechanisms. All instructions are provided in clear, professional text, and the split between static and dynamic layers is used for organizational efficiency rather than concealment.

nature-response