القائمة
Zero-ceremony single-Unit execution — run one task straight through a Station's Worker loop with no Run, no decomposition, and no state written under .darkrun/
التثبيت باستخدام Codex أو Claude انسخ هذا Prompt والصقه في Codex أو Claude أو مساعد آخر ليراجع صفحة Skill ويثبّتها لك.
استنادا إلى تصنيف SOC المهني
Start a new darkrun Run — describe what you want to accomplish and the factory manager scaffolds a right-sized lifecycle for it
Create a composite Run combining stations from multiple factories with sync points
Configure darkrun for this project — auto-detect VCS, hosting, CI/CD, and default branch, confirm with the user, and write .darkrun/settings.yml
Parking lot for ideas not yet ready to become a Run — add, list, review, or promote a backlog item into a Run
Review and decide a Station's Checkpoint — approve to advance the Run, or request changes to route rework as drift
Run a Run lights-out in dark mode — pre-elaborate the work up front, then advance Station to Station without stopping for review, pausing only on external/await gates and ambiguity
| name | darkrun-zap |
| description | Zero-ceremony single-Unit execution — run one task straight through a Station's Worker loop with no Run, no decomposition, and no state written under .darkrun/ |
Stateless single-task execution — no Run record, no decomposition, nothing written under .darkrun/. Call darkrun_zap { task, factory?, station? } and follow the returned message verbatim (it carries the Worker sequence, per-Worker subagent prompts, and the run/verify/commit procedure). If it returns a *_not_found error, surface the valid_* list via AskUserQuestion and retry.
For multi-Station or tracked work use /darkrun:darkrun-new (right-size it with --size quick for small self-contained work).
