sdlc-looplisted

# sdlc-loop Drive the backlog autonomously. The Python helpers own state/budget + the backlog source; you run each goal. The source is config-selected (`.sdlc/config.json` → `discovery.source`): **local goal files** (default) or **GitHub issues** (`source: github`, needs an authenticated `gh`). You run the loop the same way either way — the helper handles where goals come from and how status is recorded. First, reset the per-run budget: `python3 "${CLAUDE_SKILL_DIR}/scripts/loop.py" start .sdlc` (mid-session, `start-run .sdlc` resets just that budget cursor, for a deliberate fresh push without `start`'s config-warning/session-marker side effects and without hand-editing `config.json`.) Then repeat until the helper says stop: 1. `goals=$(python3 "${CLAUDE_SKILL_DIR}/scripts/loop.py" next-batch .sdlc)` — one goal per line, or a lone `DONE`/`BUDGET`. Off by default (`parallel.goals.enabled`, config-gated): with it off, or only one goal available, this is byte-identical to plain `next` — exactly one line — and everything below runs inline exactly as documented, for that one goal (`goal=$goals`). **1a. More than one line? Dispatch the batch, don't queue it** (F10.5-3/#375) — mirrors 3b's slice-wave shape one level up: instead of ONE goal's implementation slices running concurrently, MULTIPLE goals run concurrently, each all the way through its own steps 2-7. For each goal in the batch, dispatch a **subagent** (fresh context) that runs this skill's steps 2