Prompt file imported from olddognewflex/tkt (.windsurf/workflows/deploy-ready.md). Copyright stays with the author.

/deploy-ready

Pick up tickets in deploy_ready, annotate QA lane times, merge the PR, watch staging, gate manual production deploy. Ticketing via tkt; Git host via gh.

Steps

1. Find deploy_ready tickets: tkt list --query deploy_ready --json. 1.5. After-hours hold: tkt schedule --json reports label and in_window (exit 2 = misconfigured: fail closed and hold). Under tkt run KEY consider only KEY. During business hours drop tickets carrying the label and take the next eligible ticket; held tickets wait without blocking others. Only when every deploy_ready ticket is held: comment each, echo HOLD, stop before merge/staging (under tkt run, outcome gate). Re-run outside the window proceeds normally.
2. Annotate QA-lane times: tkt lane-time "$KEY" --role qa_ready, qa, deploy_ready.
3. Find the PR: gh pr list --repo $(tkt cfg vcs.repo) --search "$KEY in:title,body".
4. Merge: gh pr merge "$PR" --repo $(tkt cfg vcs.repo) --$(tkt cfg vcs.merge) --auto. Poll until MERGED.
5. Watch staging workflow matched by merge commit SHA.
6. Comment and hand off to QE.
7. Present production deploy gate (manual); if tkt schedule reports business hours and any deploy_ready ticket carries the after-hours label, warn to trigger it only outside the window.
8. Monitor production deploy once triggered.
9. Transition to done only if prod workflow truly deploys; otherwise leave in deploy_ready.

Output

• Tickets transitioned (or left for human Done)
• Production run URL
• Any deploy friction

Rules

• Repo/workflow names from tkt cfg.
• Production deploy is manual.
• tkt lane-time is a no-op when time tracking is disabled.