openspec-apply-change

Applying OpenSpec changes involves selecting the correct change, loading schema-aware context files, and iteratively executing pending implementation tasks. You can start, continue, or resume task implementation while maintaining alignment to the spec's current state.

Schema-driven workflow orchestration prompts you to select an active change, or infers it from conversation context when multiple exist. It then checks the change status to determine the schema and locate the task artifacts.

Yes, you need the openspec CLI dependency to run context-driven implementation loops. The workflow uses openspec status and openspec instructions apply with JSON parsing to retrieve context files and iterate through pending tasks.

When task implementation encounters ambiguity or blocked states, the workflow pauses safely. It stops the iterative execution loop to prevent misalignment, preserving your current spec state until the blocker is resolved.

The best way to continue implementing a partially completed OpenSpec change is to re-invoke the application workflow. It reads the current change status, identifies pending tasks, and resumes iterative execution exactly where you left off.