Sprint Planner Skill

Sprint planning automation generates disciplined plans with in_scope and out_of_scope boundaries, task dependencies, and agent configuration. Use plan-sprint.sh with your epic JSON to produce structured deliverables and iteration estimates, reducing ambiguity across orchestration waves.

Yes. Sprint execution supports Redis-backed checkpoint persistence with TTL, enabling recovery from failures at wave and task levels. Checkpoints are saved automatically, and resume-wave.sh restarts execution from the last valid state without reprocessing completed work.

Structure sprints using CFN Loop orchestration with wave and task-level checkpointing. Execute tasks via execute-sprint-task.sh, save checkpoints with save-checkpoint.sh, and clean up orphaned resources with cleanup-orphans.sh for deterministic progress across container and agent boundaries.

Yes. Sprint planning supports software-development as a task type and generates agent recommendations and context injection tailored to that domain. Specify --task-type "software-development" when invoking plan-sprint.sh to receive focused deliverables.

Yes. Sprint planning requires epic JSON as input to parse deliverables, dependencies, and scope. Pass your epic structure via the --epic-json parameter to plan-sprint.sh; the Skill transforms it into executable sprint plans and checkpoint metadata.

Checkpoint recovery relies on Redis-backed persistence with TTL. If Redis is unavailable, in-flight wave and task state cannot be persisted, preventing reliable crash recovery. Ensure Redis is accessible and TTL is configured before executing multi-wave orchestrations.