Align

Use $align to equalize context before action. The skill exists for tasks where the human, project agent, WebGPT, $scillm, $ask, $dogpile, $memory, or $project-knowledge may each know different facts. It produces an alignment_lock.json that later skills can execute from.

Do not use $align to implement, render, refactor, or review final work. It is a pre-execution gate. No code, infographic, dashboard, plan-iterate phase, or project-infographic render should start until the alignment lock says ready_for_execution: true.

Core Rule

Alignment is complete only when all blocking questions are answered and the lock records: