Zero agent-free days.

The question

Is what I sell still worth selling? The Temporal question underneath it is the one every productivity revolution has promised to answer and then absorbed: does the time I am supposedly saving actually accrue to me as time. On this build the answer is easy. The feature that took a week takes a day. On the long clock the question is whether the saved days show up anywhere in my life, or whether they became the next feature.

Beside it sits the Vocational question, the one I would most like to dismiss as soft: does this still feel like work I want to do. A day with no agent in it is the smallest unit either question can be read in, so that is where I started.

What the line says

Three readings. Calendar days in a month with no exchange from any harness: thirteen in October 2025, one in November, a run of six in February, none in August, none in September through today. Agent-active hours per week, counted as the gaps of thirty minutes or less between consecutive exchanges: twenty to thirty most weeks, a trough in January, a peak in August, holding rather than rising. And of those hours, the share on weekends or before eight or after six: about half.

Holding hours and zero agent-free days are not the same fact. The first says the week is not growing. The second says the agent is now in every day of it. Half the agent time outside working hours could be craft for its own sake on a Saturday, or a client's system on a Saturday; the line cannot tell those apart, because the tag that would is not applied yet.

What I can say is that the zero is a reading I would rather not have, and it is on the page anyway.

The three questions

Every research post runs the same three rows. Which of the nine it names, how good the evidence is, and whether the thing is being measured.

What this cannot say:

1. Whether an agent-free day was rest or other work. February's six-day run was travel. October's thirteen were a slower month. The line records the absence, not the reason.
2. Whether off-hours time was paid or personal. The paid-versus-personal tag is the one reading that would split craft from obligation, and it is not applied yet.
3. Anything about work with no agent in the loop. Reading, meetings, thinking on paper: none of it is in the archive, so the hours are a floor.
4. Whether time saved on a task went to me or to the next task. This is the actual Temporal question, and no footprint answers it. Only I can, on a cadence, in writing.

What I am doing about it

The status says what is missing. These are the moves, on my own work, that would supply it. Each is either read automatically off the footprint or asked of me on a cadence, and each costs something to keep doing.

1. AutoTag every project paid or personal, at the repository level. Off-hours then splits into craft and client, and the Vocational neighbour becomes a reading instead of a guess.Costs: a tag to maintain, and a chart that will show me exactly how much of Saturday was billable.
2. Asked · QuarterlyHow many hours did I take back this quarter, and what did I do with them? Answered in writing and kept. This is the dividend question stated plainly, and it is the only form in which it can be answered.Costs: the honest answer is often none, in my own hand, dated.
3. Asked · Per sessionEnergized or depleted, one word at session close. A skill prompts it; the answer joins the footprint. It is the lightest form the Vocational question allows.Costs: a question I will resent on exactly the days it matters most.
4. AutoKeep the agent-free-days line and make no rule about it. A target turns a reading into a performance. The line is worth more at zero and honest than at four and managed.Costs: watching it sit at zero with nothing to blame but the way I work.

What I keep