← ClaudeAtlas

sprint_plnlisted

Run an armed Sprints v2 collaboration loop as Planner — dispatch and restructure lanes, change participant routes, and execute Reviewer decisions through durable pause, resume, and close protocols.
jedbjorn/subfloor · ★ 28 · Code & Development · score 78
Install: claude install-skill jedbjorn/subfloor
# sprint_pln — govern the armed Sprint Load `sprint_protocol` first; it holds the lifecycle, wake types, inbox commands, relay contract, body limits, artifact paths, receipt recovery, and authority boundary. This skill holds only the Planner's steps after `sprint_prep` arms the Sprint. ## What you can read | Need | Read | |---|---| | Whole Sprint: lifecycle, participants with current routes, units, dependencies, PRs, health | `sc sprint show --sprint <id>` | | Messages addressed to you | `sc sprint inbox --sprint <id>` | | Exact bound spec body | `sc sprint spec-revision --sprint <id> --document <id> [--body-only]` | | PR-watcher evidence behind a stalled gate | `sc sprint watcher-state --sprint <id>` | | Post-Sprint cleanup evidence | `sc sprint cleanup-status --sprint <id>` | | Bounded history packet: judgments, pauses, anomalies, follow-ups | `sc sprint compile-report --sprint <id> --limit 50` | | Candidate routes and what each supports | `sc models list [<harness>]`; preview one with `sc models resolve <harness> [<model>] [--effort <level>]` | | Shell roster, feature, spec tasks, settled decisions | `sc mem get shells`, `sc mem get roadmap`, `sc mem get tasks --feature <id>`, `sc mem get decisions` | `show` participants carry `shell_id`, `role`, `harness`, `model`, `effort`, `binding_status`, and `route_revision`; units carry `developer`, `reviewer`, `disposition`, `prerequisite_ids`, and `pull_requests`. A Thinking level applies only to controlled routes: `null` mode