sprint

Multi-agent orchestration for sprint execution coordinates parallel AI sub-agents to decompose and implement epics end-to-end. It routes tickets through sub-skills for preplanning and implementation planning, applying safety gates to prevent unsafe work.

Automating epic decomposition uses an orchestrator to route epic-typed tickets through preplanning and implementation-plan sub-skills. This structured routing automates sprint workflows by delegating ticket analysis and validation to parallel sub-agents.

Drift detection during sprint preplanning works by applying gatekeeper safety checks, including clue gates and clarity gates. If drift or unclear work is detected, it triggers a replan cascade to prevent unsafe execution.

Yes, the multi-agent orchestrator applies across epic, bug, and story tasks. It routes non-epic tickets through appropriate sub-skills like brainstorm, while maintaining context management and gating for safe automation.

No, you do not need to manually trigger sub-agents. To start sprint orchestration, you select a primary ticket id and let the orchestrator coordinate all sub-agents end-to-end automatically.

Cascade replan happens in automated sprint workflows when gatekeeper safety checks detect drift, missing clues, or unclear task definitions. This prevents ambiguous or unsafe work by forcing a structured re-evaluation of the implementation plan.