executing-plans

Sequential plan execution is achieved by processing plan tasks one by one with guardrails that enforce non-disruptive changes, readability checks, and explicit verification steps before reporting results. It pauses to clarify ambiguities whenever needed.

Plan continuity allows you to resume execution from where a plan left off by simply requesting continued execution. The skill identifies the remaining tasks in your plan file or inline plan and proceeds sequentially from the current starting point.

Yes, you can execute tasks from an inline plan or a dedicated plan file such as docs/plans/*.md. The skill applies the same sequential execution, guardrails, and verification steps regardless of the plan source format.

When a task is unclear during plan execution, the workflow pauses and asks a single precise question to clarify the ambiguity. This guarded decision-making approach prevents incorrect execution and ensures non-disruptive changes to your documentation.

No specific dependencies or components are required to start executing plans with guardrails. You only need an existing plan file or inline plan to begin sequential task execution with built-in readability and verification checks.

You should avoid automated plan execution when tasks lack clear sequential steps or when your documentation requires uncontrolled structural changes. The workflow enforces non-disruptive changes and readability preservation, limiting free-form or highly ambiguous modifications.