Kanban board management for GitLab issues with AI automation
Invoke this skill in Claude Code:
| Command | Description |
|---|---|
/gitlab-board cicd |
Show board status for cicd project |
/gitlab-board cicd all |
Work on all ready items in cicd project |
/gitlab-board cicd 22 |
Work on issue #22 specifically |
/gitlab-board cicd 22,23,24 |
Work on issues #22, #23, and #24 |
/gitlab-board cicd ready |
Work on all items in ready column |
/gitlab-board cicd review |
Show items in review column |
Meta-cards that spawn task cards when activated. Drag to ready and AI creates all tasks for that phase. (v1.11.0+)
Grey area choices requiring human input. Uses checkbox voting - human checks an option, AI finalizes the decision.
Reference documentation and summaries. Not actionable work - for context and onboarding.
Work items that flow through development. Includes bugs, enhancements, and infrastructure tasks.
Move issue from ready to in-progress, signaling active work.
Parse issue description and comments to understand the task.
Create feature branch: issue-{IID}-{description}
Edit files, run commands, implement the requested changes.
Stage changes, commit with issue reference, push to remote.
Create merge request linking back to the issue.
Update issue to review status, add completion comment.
Human reviews MR, approves/requests changes, merges.
### Options (check ONE to decide)
- [ ] **A**: Option A description
- [x] **B**: Option B description (SELECTED)
- [ ] **C**: Option C description
- [ ] **Other**: See comment below
Review options, check the preferred checkbox, move card to "ready".
Detect checked option, record decision, add comment, close issue.