dispatching-spec-reviewlisted

# Dispatching a Spec Review **Render before asking.** Before any AskUserQuestion, render the decision context as assistant chat text. Content that exists only in a Read result, a subagent/Agent return, or the question/option strings **does not count as rendered** (the UI shows a compact card) — restate it in chat first. Sends a fresh-context subagent to compare what the spec requires against what the diff delivers. The subagent reads, reasons, and reports — it does not edit code or commit. The orchestrator uses the structured findings to decide whether to mark a task done, request fixes, or block approval. ## When to Invoke | Caller | Scope | Trigger | |---|---|---| | `/ship-execute` post-task gate | One task | After `dispatching-task-loop` returns `STATUS: COMPLETE` and anti-stub-scan is clean — final compliance check before marking task done | | `/ship-execute` wave VERIFY | All tasks in a wave | After wave-boundary REFACTOR — confirms wave-level acceptance scenarios all hold | | `/ship-review` | Sprint or feature | Full audit before user approval | | Manual / ad-hoc | Single feature | When the user asks "did we actually deliver F-007?" | Skip this skill for tasks marked `effort: S` (trivial) — overhead exceeds value. The post-task gate explicitly bypasses for S tasks per `/ship-execute`'s spec-check rule. ## Inputs - `scope` — `"task" | "wave" | "feature" | "sprint"`. Determines which spec files and which diff range. - `target_ids` — list of task IDs (scope=task), f