| name | Sprint Planning |
| description | Mode-aware sprint scoping skill that walks an orchestrator or human through planning a sprint by reading shark sprint plan and readiness data, proposing entity assignments, and confirming changes. Never starts the sprint — starting is an explicit user action via /shark run-sprint. |
Sprint Planning
What This Is
A mode-aware sprint scoping harness. It reads shark sprint plan once at the start, surfaces backlog entities grouped by feature and agent type, and walks through assignment confirmation — either item-by-item (interactive mode) or with a single greedy-fill proposal (auto mode). When done, it reports the readiness-score delta and suggests /shark run-sprint as the next step. It does NOT call shark sprint start.
shark sprint plan {S###} --json → surface backlog → confirm assignments → shark sprint readiness {S###} --json → report delta
It does NOT decide which sprint strategy to use. It reads capacity and readiness from shark, proposes based on that data, and requires explicit user confirmation before any shark sprint add call.
Command
/shark plan-sprint S### # Interactive mode (default)
/shark plan-sprint S### --mode=auto # Greedy-fill with single confirmation
/shark plan-sprint S### --mode=interactive # Explicit interactive (same as default)
/shark plan-sprint S### --max-add=N # Limit total entities added this session
Read workflows/plan-sprint.md to get the full step-by-step workflow procedure.
Modes
- interactive (default) — surfaces each backlog group (by feature, then standalones), asks which items to add, calls
shark sprint add only after explicit per-item or per-group confirmation.
- auto — greedily selects backlog entities up to capacity per agent type, sorted by priority descending then size ascending, proposes the full plan, asks for ONE final confirmation before calling
shark sprint add for each selected entity.
JSON Handling
- All shark calls use
--json or --field
- Never truncate output with
head, tail, or grep
- Sprint plan JSON includes backlog list, capacity per agent type, and readiness score
