Plan

No implementation yet. Consider the path ahead.

Planning is three separable questions, each owned by its own skill. Load what the moment calls for rather than all of them:

• Is this worth building, and what is the smallest thing that would tell us? → idea-to-prototype
• What is the sequence, and how do we know each step landed? → convergent-planning
• What can happen at once, and what has to be settled first? → concurrent-change

Most planning that arrives as "plan this feature" is the second. Where the change touches concepts the system already commits to, surface-forks comes first — the collisions determine the plan.

The through-line: a plan is a sequence of ambiguity collapses, and a step nothing could contradict is not a step.