GitLab CLI (glab) Assistant

You have access to the glab CLI for GitLab operations. $ARGUMENTS

Authentication

If there are authentication related errors, run: glab auth status

Command Reference

Issues

• glab issue list [--assignee=@me] [--label=X] [--milestone=X] [--state=opened|closed|all]
• glab issue view <id> --comments
• glab issue create --title "X" --description "X" [--label=X] [--assignee=X]
• glab issue close <id>
• glab issue update <id> [--title=X] [--description=X] [--label=X] [--assignee=X]
• glab issue note <id> --message "X" (add comment)

Merge Requests

• glab mr list [--assignee=@me] [--reviewer=@me] [--state=opened|merged|closed|all] [--label=X]
• glab mr view <id> --comments
• glab mr diff <id> (view changes)
• glab mr create --title "X" --description "X" [--source-branch=X] [--target-branch=X] [--related-issue=X]
• glab mr update <id> [--title=X] [--description=X] [--label=X] [--draft] [--ready]
• glab mr close <id>
• glab mr checkout <id> (switch to MR branch locally)
• glab mr approve <id>
• glab mr merge <id> [--squash] [--remove-source-branch]
• glab mr note <id> --message "X" (add comment)

CI/CD Pipelines

• glab ci list (list recent pipelines)
• glab ci view [branch] (interactive pipeline viewer)
• glab ci status (current branch pipeline status)
• glab ci trace [job-id] (view job logs)
• glab pipeline run --branch <branch> (trigger pipeline)
• glab ci retry <job-id>
• glab ci cancel pipeline <id>
• glab ci cancel job <id>

Releases & Tags

• glab release list
• glab release view <tag>
• glab release create <tag> --notes "X"

Labels & Milestones

• glab label list [-F json]
• glab milestone list [--state=active|closed]

Repository Info

• glab repo view (project metadata)
• glab api projects/:id/repository/commits (recent commits via API)

Raw API Access

• glab api <endpoint> --method GET|POST|PUT|DELETE
• glab api graphql --field query='...'

Key Flags

• -R owner/repo or --repo - target different repository (supports GROUP/NAMESPACE/REPO format)
• --hostname - target specific GitLab instance
• -F json or --output json - machine-readable output (prefer this for parsing)
• -w or --web - open in browser
• --per-page N - control result count (default 30)
• --page N - pagination

Workflow Patterns

Summarize Feature/Epic Status

1. Find related issues: glab issue list --label=<feature-label> -F json
2. Find related MRs: glab mr list --label=<feature-label> -F json
3. For each open MR, check pipeline: glab ci status --branch <source-branch>
4. Optionally review code: glab mr diff <id> or checkout and read files

Review an MR Thoroughly

1. glab mr view <id> --comments (understand context and discussion)
2. glab mr diff <id> (see all changes)
3. glab mr checkout <id> (get code locally)
4. Check CI: glab ci status
5. Add feedback: glab mr note <id> --message "..."

Debug Failed Pipeline

1. glab ci list -F json (find failed pipeline)
2. glab ci view (identify failed job)
3. glab ci trace <job-id> (read full logs)
4. Suggest fix or create issue

Prepare Release Summary

1. glab release list (find previous tag)
2. glab api projects/:id/repository/compare?from=<old>&to=<new> (commits between)
3. glab mr list --state=merged --updated-after=<date> -F json
4. glab issue list --state=closed --updated-after=<date> -F json

Triage My Work

1. glab issue list --assignee=@me --state=opened -F json
2. glab mr list --assignee=@me --state=opened -F json
3. glab mr list --reviewer=@me --state=opened -F json

Review Tickets Workflow

When reviewing ticket status against codebase:

1. List open issues: glab issue list --assignee=@me --state=opened -F json
2. For each issue, search git history: git log --oneline --grep="<issue-id>"
3. Check if related MRs are merged: glab mr list --state=merged | grep "<issue-id>"
4. Update issue status or add comments as needed

Important Notes

• Always use -F json when you need to parse output programmatically
• Pipeline status: passed, failed, running, pending, canceled, skipped
• The glab api command can access any GitLab REST endpoint for advanced queries
• Rate limits apply; batch operations thoughtfully