/omni:mega-drive
drive PRDs across repositories, one step per tick
What it does
The /omni:drive of a plan repository: run under /loop, it drives your own multi-repository PRDs, the ones you name, or with --roadmap <n> exactly roadmap n's. Its loop plan puts two steps in series only when they touch the same path in the same repository. Each tick runs one skill (/omni:ultra-wave, /omni:ultra-yolo, which plans a PRD with no plan, /omni:ultra-yolo-fix or /omni:mega-pr-care --once) or waits, parks a PRD waiting on a person on its plan PR's status comment naming each open PR by repository, records the repositories the tick touched on the Loop page, and under --roadmap sends the roadmap's page. Outside a plan repository it prints the /omni:drive line and stops. It never answers the outbox and never merges into any default branch.
When to use it
Use it when multi-repository PRDs are merged into the inbox of a plan repository and you want them built in every target without typing each next command.
How to use it
/loop /omni:mega-drive [<n>…]
/loop /omni:mega-drive --roadmap <n>Example
/omni:mega-drive→ one tick across the repositories; under /loop, every tick until your PRDs wait on you
Who runs it
You type it in Claude Code.
Related skills
Its full instructions, written for Claude: kit/plugin/skills/mega-drive/SKILL.md