Rendered from docs/obligations/0112-a-surface-cannot-move-the-assisted-fraction-of-a-run.md in the Headwater corpus. Every document on this half of the site is typed by the taxonomy the descriptor names: corpus.json.

A surface cannot move the assisted fraction of a run

Context

Q7 claims that a working-tree write tool raises the assisted fraction. HW-OBL-0004 makes that claim a comparison of a baseline arm and a treatment arm. The capture-cost store now names the arm of every reading, and headwater capture groups by it.

Obligation

The four counts of a reading are derived from the plan. The scaffolder proposes a plan from a kind, a title, a date and the relations a caller named. The surface is not one of those inputs, and no field of a plan carries one.

So two runs of one kind, with the same relations, write the same four counts whatever asked for them. A comparison of the per-run fraction across two arms therefore measures nothing about either arm. What a surface can move is the count of runs, which is the reach figure, and the mix of kinds that those runs chose. It cannot move the number inside one run.

Discharge

Measured on 2026-08-14 over this repository. Two readings of one kind, one from each arm, carry the same four pairs. Front matter is 4 of 5, sections 3 of 3, identifier 1 of 1, and edge halves 0 of 0. Each run is 8 of 9. The two lines of the store differ in surface, in document and in id, and in nothing else.

The comparison that HW-OBL-0004 asks for therefore reads one of two other things. The first is reach, which is a count over runs that happened. The report already states it. The second is the aggregate fraction over a mix of kinds that the two surfaces reached for differently. That second reading is a claim about what an author picks. It needs many runs and a stated population. This corpus holds five readings, three of which name an arm, so neither is available.

This record is discharged. The Obligation section states a fact about the scaffolder, and the measurement above proves that fact. So this record is a finding, and it owes nothing more. HW-OBL-0004 stays open as the Q7 claim and carries the comparison. The store held 302 readings on 2026-09-30: 299 named the terminal surface, 1 named the protocol surface, and 2 named none. The owner ruled this on 2026-09-30, in the roadmap review of that day.