openspec-apply-change

To implement OpenSpec changes, this skill resolves the target change from context, reads workflow state, and executes pending tasks while updating task checkboxes to track progress. It requires the openspec CLI to load context files and manage the change-driven implementation session.

The best way to resume an OpenSpec change is to let the skill read the workflow state and apply instructions to determine current progress. It identifies remaining pending tasks and continues executing them one by one, making minimal code changes as needed.

OpenSpec change management handles ambiguity by triggering safe pausing. The skill stops cleanly when information is missing, a task is ambiguous, or a blocker appears, then tells the user exactly what to do next to resolve the issue.

Yes, you need the openspec CLI to execute spec-driven change tasks. The CLI is required to load context files, track task progress, update checkboxes, and stop safely when blockers or missing artifacts occur during the implementation workflow.

Yes, you can apply a specific OpenSpec change by name. The change selection feature resolves the target change from a provided name, conversation context, or the list of active changes, automatically determining the required schema and context files.

Automation is limited by safe pausing constraints; it stops when information is missing, tasks are ambiguous, or blockers appear. It also requires existing workflow state and artifacts, making minimal code changes rather than handling full architectural implementations.