Something you were doing steadily stops. Nothing dramatic happens; the sessions just thin out and then stop, and a month later it is on the list of things you are vaguely ashamed of. The default response is to apply more force. That is the right response to exactly one of the three common causes, and it actively worsens the other two.
So the first move is diagnosis, and the diagnosis is usually quick if you ask the right question.
Cause one: the next action is not defined
This is the most common and the least interesting. The project has reached a point where the next step is genuinely unclear — not difficult, unclear — and each time you sit down you spend the available energy working out what to do and none of it doing anything. After a few rounds of that, sitting down starts to feel unrewarding, and you stop.
The test: can you say, in one concrete sentence, the physical thing you would do first if you sat down right now? If the answer needs an "and then I would work out whether", this is your cause.
The response: stop trying to advance the project and spend one short session defining the next three actions, written concretely enough that a stranger could do them. This is the one case where the stall dissolves as soon as it is named.
Cause two: there is a cost you have not priced
The work requires something you have not admitted it requires, and the unadmitted thing is doing the stopping. Common examples: it needs a conversation you are dreading; it will expose that the earlier version was wrong; finishing it means the next thing has to start; doing it well requires learning something that will make you feel like a beginner again.
The test: imagine the project completed successfully next Friday. If there is any flicker of relief at the thought of it not happening, or any specific consequence you would rather not face, this is your cause.
The response: name the cost out loud or on paper, in one sentence, without softening it. A surprising number of stalls end there, because an unnamed cost operates as generalised reluctance and a named one turns into a task — usually a single uncomfortable conversation, which can be scheduled like anything else.
An unnamed cost is felt as a mood. A named one is a task with a date, and tasks are much easier to do than moods are to overcome.
Cause three: you no longer want it
The goal was set by someone with different information, or it was never really yours — it belonged to a job you have left, a comparison you were making at the time, or an expectation you had absorbed. The stall is not a malfunction. It is an accurate response.
The test: if the goal were completed for you overnight, at no cost, would you be pleased or merely relieved? Relief is the tell. Pleasure survives the fantasy; obligation does not.
The response: stop, deliberately and on the record, rather than by attrition. The cost of dropping something explicitly is one uncomfortable afternoon. The cost of dropping it by attrition is months of low-grade guilt plus the erosion of your own belief that you finish things. If you cannot bring yourself to stop, park it with a review date; that is at least an honest state.
Do not skip to the third cause
Cause three is unusually attractive on a bad day, because abandoning something can be dressed as self-knowledge. Run the first two tests first, in order. If the next action is undefined, define it and do one session before deciding anything about wanting. Judgements about desire made while stuck are unreliable, in both directions.
Restarting
Whichever cause it was, restarting has the same shape. Come back at half size or less, at the original fixture if there was one, and do not attempt to make up lost ground — the impulse to compensate is what produces the second collapse a fortnight later. One session at reduced scale, then another. The point of the restart is to re-establish the sequence, not to recover the calendar.
And keep the diagnosis. If the same cause turns up on the next three things you stall on, that is worth more than any of the individual fixes: it tells you which of the three is your characteristic failure, and you can design against it in advance next time.