gh:work

You can automate work execution by providing a plan document or bare prompt to initiate task decomposition and phased planning. This process yields a shipped feature by applying memory-backed context retrieval, quality checks, and incremental commits.

Task decomposition is the process of breaking a plan document or bare prompt into structured, trackable tasks for guided activities. It enables phased planning, memory context retrieval, and quality checks across iterations and subagents.

Yes, memory-backed context retrieval provides historical session awareness to reuse prior work patterns during work execution. It maintains memory context across iterations and subagents to satisfy traceable task requirements and execution contracts.

Yes, you can execute a bare prompt to ship complete features by transforming it into structured, trackable tasks. The automation applies phase-based guardrails, integrated quality checks, and incremental commits to yield the shipped feature.

The best way to ensure quality checks across iterations is to apply phase-based guardrails with memory-backed context retrieval during work execution. This integrates testing and incremental commits, satisfying verification criteria aligned with an established harness.

Yes, you need an established harness to apply traceable task execution contracts. The automation satisfies requirements for traceable tasks and verification criteria aligned with this harness to ship complete features successfully.