Loop Skill

Run another agent skill multiple times in sequence, reporting only the final net result.

Arguments

• Target skill invocation: required. Accept $skill-name, skill-name, a path to SKILL.md, or a natural-language instruction that names the skill and its arguments.
• --runs <n> or -n <n>: optional positive integer iteration count. Default: 3.
• Additional target arguments: forward unchanged to the target skill on every iteration.

Preconditions