subagent-driven-development

To execute implementation plans with isolated context, this Skill dispatches a fresh subagent per task, preventing cross-task context leakage and enforcing per-task isolation for safer iterative sessions.

Running multi-task plans in parallel without context leakage is achieved by dispatching dedicated subagents. Each task gets isolated context, preventing interference during parallel execution.

To enforce code quality reviews during automated workflow execution, the system applies a two-stage review loop after each task: checking spec compliance first, then code quality before handoffs.

Yes, you can run sequential task execution with controlled handoffs between subagents. The Skill orchestrates staged reviews and manages self-review and automated handoffs between tasks to ensure traceability.

Parallel execution of tasks requires fresh subagents for each iteration to maintain per-task isolation. This approach accelerates iterative feedback loops by preventing cross-task context bleed.

Cross-task context leakage happens in iterative development sessions when a single context window accumulates overlapping task data. Dispatching fresh subagents per task isolates context and prevents this bleed.