The Agent Skills Directory

[COMMAND_EXECUTION]: The skill instructions guide the agent to include executable verification methods, such as terminal commands, within plan tasks and completion checklists to confirm task success.
[SAFE]: The skill implements a localized file management workflow, saving and archiving plan documents within the ./docs/plans/ directory of the repository. These actions are restricted to the project environment and do not exhibit malicious behaviors like data exfiltration or unauthorized access.

writing-plans