---
name: review-queue
description: Triage and review every open PR in a repo — Jev picks depth and model, code picks merge order
argument-hint: "OWNER/REPO (default: this repo's origin)"
---

Load the `qa` skill, REVIEW mode. Target: $ARGUMENTS (else `gh repo view --json nameWithOwner`).

1. Run `python3 ~/.claude/skills/qa/review.py <repo>`; show the table as-is.
2. `covered` PRs: list them with the PR that contains them; offer to close (ask first).
3. `blocked`: name the failing check per PR (`gh pr checks`), don't review.
4. For the rest, in merge order, launch one review agent per PR in ONE message, `model` from the
   table; `deep` agents start from the hot files and add a `qa` PLAN of the blast radius.
5. Report per PR: verdict (approve / changes / block), findings with file:line, merge order. No
   approvals, merges or closes without the user saying so.

Close with (these were the follow-ups every studied run of this command got): `acted X of N`
over the leads, each one test written / fixed (commit) / issue (URL) / dismissed with a reason;
the live state of any background agent or workflow, read now rather than from memory ("are the
workflows running though or not???"); and, when work continues in another session, /wrap step
8's continuation prompt. Nothing is "on main" until `~/.claude/commands/bin/repo-hygiene.sh --landed` says so.