wave-planlisted

# wave-plan Answer "which eligible issues could form the next wave, and can they run alongside what's already in flight?" This is the strategic layer — read-only, advisory, and the only output is a report. Materializing one approved wave into a spine and worktrees is `wave-create`'s job. Your job is the **judgment** — reading the report, flagging what needs attention, and helping the human pick the right set of ids. The CLI plumbing (the four commands, the exact sequence, the worked `CrossWaveResult` sample) lives in [reference/plan-mechanics.md](reference/plan-mechanics.md) — reach for it once you need to drive the engine. You never touch a tracker directly; everything goes through the engine CLI (`{{wave-cli}}`), which selects the configured store. **M1 scope: candidate draw + cross-wave report + PRD panel + the read-only goal panel + the dispatch-cost estimate. No batch-sizing heuristics** — sizing by worker-mix, risk-mix, or wallclock estimates is a soft CHARTER idea, not an M1 requirement. Do not invent heuristics. The dispatch-cost estimate (step 3c) is **not** an exception to that rule: it is arithmetic over the set in front of you, reported so the human can see the bill. It scores nothing, ranks nothing, filters nothing, and recommends no size. ## When to Use - Planning the next wave from the backlog: you need to know which issues are eligible and whether running them beside current claims is safe. - Running a cross-wave check before handing ids to `wave-create`.