implement-worktree-no-merge

To implement code changes in an isolated Git worktree without merging, this skill creates a new branch from the current HEAD, sets up the worktree, and commits changes phase by phase. It leaves the worktree intact for an orchestrator to handle testing and merging independently.

The best way to commit code phase by phase in a separate Git branch is using a worktree-based implementation skill. It performs deep system understanding via subagents before applying changes, ensuring phased commits are contained within the isolated branch for orchestrator review.

Yes, an orchestrator can manage testing and merging separately after isolated implementation. The skill emits worktree and branch details upon completion, leaving the isolated environment intact so the orchestrator can independently execute the testing and merging workflows.

Git worktree setup handles deep system understanding by deploying subagents to analyze the system before implementing code changes. After creating a new branch from the current HEAD, subagents perform deep analysis to ensure accurate phased commits within the isolated worktree.

You need an orchestrator to manage the downstream testing and merging processes when using this isolated worktree skill. The skill focuses strictly on implementation and phased commits, leaving the worktree intact specifically for an orchestrator to handle subsequent integration.

The limitation of implementing code in an isolated worktree without merging is that the skill does not execute testing or merging. It strictly handles branch creation, worktree setup, deep system understanding, and phased code commits, requiring a separate orchestrator for final integration.