The Agent Skills Directory

[SAFE]: The skill functions as a passive dispatcher and orchestration layer, delegating actual task implementation to other specific modules.
[SAFE]: Includes explicit safety constraints, such as 'Hard Caps' limiting the number of iteration loops (max 5) and tool calls (max 50), which mitigates risks of unintended autonomous behavior.
[SAFE]: Requires human-in-the-loop validation for multi-skill execution chains and explicitly forbids automated Git operations (commit, push, merge) without direct user authorization.

klh-dispatch