executing-plans

Executing a development plan step-by-step involves loading a plan file, reviewing it critically, and converting it into a sequence of tracked tasks. The process applies verifications after each task and updates progress states like in_progress and completed until the workflow finishes.

The best way to review a plan before task execution is to load the plan file and critically evaluate its steps. This mechanism converts the reviewed plan into executable tasks, ensuring verifications are defined and blockers are identified before development begins.

Handling blockers during plan-driven workflow execution requires escalating them when a step cannot be completed. The execution mechanism pauses the task sequence, reports the blocker, and prevents state updates to completed until the issue is resolved.

Yes, you can use required sub-skills to finalize development after executing tasks. Sub-skills like finishing-a-development-branch integrate with the execution workflow to complete the operational sequence and present the final results.

Yes, you need a pre-loaded plan file to start task execution and verifications. The execution mechanism requires an existing plan to review critically, convert into tasks, and apply step-by-step verifications with state updates.

Workflow execution gets blocked during state updates when verifications fail or required sub-skills are missing. The execution mechanism prevents advancing task states from in_progress to completed until all defined verifications pass and blockers are explicitly escalated.