A new person arrives on a Monday. She has read the file. She knows the goals, the schedule, the targets, the data sheets. She is good at her job and she wants to do it well.
She does not know that he will not start anything while his sleeves are wet. She does not know that the word "ready" means something different to him than it does to her, or that the third time you ask is the one that lands, or that the way to get him through a door is to stop looking at the door. Nobody wrote those down. They were not in the plan. They were in a person, and that person is gone.
Families describe this part the same way to us, and they have been describing it for sixteen years. The weeks after a change are harder than the weeks before it. Things that were settled come loose. A child who had stopped protesting a routine protests it again. Progress that took a season to build is spent re-earning ground that was already won, and the adults around him call it a setback, and write it down as one.
It is worth being precise about what actually happened there, because the record will outlive everybody in the room. Nothing about the child changed on that Monday. What changed is that the only copy of a large, unwritten body of knowledge about him left the building. He is being asked to rebuild it, in the one currency he has, which is his own patience.
This is the part that is fixable, and it is not fixable by asking people to stay forever. People leave jobs. They will keep leaving jobs. The question is what they are permitted to take with them.
Most of what makes support work is not the protocol. It is the accumulated, specific, ordinary knowledge of one person — what helps, what backfires, what a particular silence means, which version of an instruction he has actually understood for the last two years. When that lives only in someone's head, it is one resignation away from gone. When it lives in the record, it is inherited.
That is a design choice, and most teams have never been asked to make it deliberately. It costs almost nothing to write down that he needs the third ask. It costs a year of a child's life to discover it again from scratch.
I do not know how much of it can be written down. Some of what a good person knows about a child is probably not reducible to a sentence, and anyone who tells you the whole of it fits in a form is selling you the form. But "some of it cannot be captured" has become the reason to capture none of it, and that is a much bigger claim than the evidence supports.
So the question worth putting to any team supporting your child, in any setting, is not only what the goals are. It is: what do you know about him that is not written anywhere? And then: could we write it down?
Ask it while the person who knows is still there. That is the whole window.
A child should not have to teach each new adult who he is. He has already done that work. The record is supposed to remember so he does not have to — at six, at nineteen, and at every handover in between.