Menu
Context report + quiz on a change (PR, branch, or session) that the maintainer must pass before merging.
Install with Codex or Claude Copy this prompt, paste it into Codex, Claude, or another assistant, and let it review the skill page and install it for you.
Based on SOC occupation classification
Run one scheduled cloud-routine fire: pick the oldest open `ready-for-agent` issue, drive it to a merge-ready PR, and STOP at the merge gate without merging. Composes `ship`. Use only when running the scheduled cloud ship routine for Gharib89/crm (the routine prompt invokes this skill by name); not for an interactive `/ship`.
Review inbound agent-shipped PRs at the merge gate — drift checklist, targeted live e2e, fix-in-place, review-bot iteration (Copilot + CodeRabbit) — leaving each PR merge-confident (gate-passed) or explicitly failed (gate-failed).
Take a tracker issue to a merge-ready PR in one unattended run, stopping only at a human merge gate. Composes the `tdd` and `code-review` skills. Use when the user wants to ship an issue or take an issue through to a PR; also invoked by the cloud ship routine.
Ask which skill or flow fits your situation. A router over the skills in this repo.
Review the changes since a fixed point (commit, branch, tag, or merge-base) along two axes — Standards (does the code follow this repo's documented coding standards?) and Spec (does the code match what the originating issue/PRD asked for?). Runs both reviews in parallel sub-agents and reports them side by side. Use when the user wants to review a branch, a PR, work-in-progress changes, or asks to "review since X".
Shared vocabulary for designing deep modules. Use when the user wants to design or improve a module's interface, find deepening opportunities, decide where a seam goes, make code more testable or AI-navigable, or when another skill needs the deep-module vocabulary.
| name | quiz-before-merge |
| description | Context report + quiz on a change (PR, branch, or session) that the maintainer must pass before merging. |
| disable-model-invocation | true |
| argument-hint | [pr-number | branch | session] |
| metadata | {"internal":true} |
An agent-built change can land with the maintainer understanding less of it than they think — the diff alone hides behavior that depends on existing code paths. This skill closes that gap: a context report, then a quiz; the merge waits until the quiz is passed.
$ARGUMENTS names a PR number, a branch, or session (the work done in this conversation). Read the full diff and the surrounding code of every touched region — call sites, shared state, error paths — because the quiz must cover how the change interacts with what was already there. For a PR, also read the body, linked issue, review threads, and any deviations section. Done when you can explain each hunk's purpose without re-reading it.
Write for a maintainer catching up, not a changelog. Cover, in order: the intuition (what changed and why, in plain language), the mechanics (how it works, which existing paths it touches), the judgment calls the agent made on its own (deviations, chosen trade-offs), and what to watch after merge. Keep it short enough to read before the quiz — the quiz enforces the reading.
Five to eight questions, answerable from the report plus reasoning — no diff trivia. Weight toward:
Ask all questions at once, wait for answers, grade honestly. For each miss, explain the correct answer, then pose one variant question on the same point.
Passed = every question, including variants, answered correctly. State pass or fail plainly. On fail, recommend against merging yet and name exactly what to re-read — the specific report section or code region, not "review the change".
