The Agent Skills Directory

[SAFE]: The skill is a configuration authoring tool that assists users in defining workflow steps and parameters.
[SAFE]: Implements security checks to prevent path traversal attacks by rejecting '..' in artifact paths and iteration sources, as documented in the schema reference.
[COMMAND_EXECUTION]: The skill allows users to define 'shell-command' verification policies. This is a core functional feature of the GSD workflow engine for validating outputs (e.g., checking if a file exists or matches a pattern).

create-workflow