Rendered from docs/obligations/0012-the-human-half-of-the-public-presence-claim-has-no-instrument.md in the Headwater corpus. Every document on this half of the site is typed by the taxonomy the descriptor names: corpus.json.

The human half of the public presence claim has no instrument at all

Context

Q16 settles the public presence of the project. A generated site should let a reader answer "is this for me?" without reading the specification.

Obligation

The Discovery probe category is the instrument for the machine half.

Discharge

The human half has no instrument at all, which is worth a statement rather than a silence. No probe category grades what a person understood from a page, and this project has proposed none.

The machine half has a reading, and it does not separate the arms. The batch of 2026-09-30 graded the two discovery probes in both arms of the campaign tier, 30 sessions of each probe in each arm (#1384). The present arm satisfied 9 of 60 (15.0%, 8.1% to 26.1%). The absent arm satisfied 5 of 60 (8.3%, 3.6% to 18.1%). The difference is +6.7 points, in a 95% Newcombe interval of -5.3 to +18.7 points (the result). The interval contains zero, so this batch does not show that the governance layer helps a cold session find a document.

The reading has limits. 5 of the 60 present-arm sessions and 3 of the 60 absent-arm sessions reached this repository from outside their workspace. That count uses the rule in the Limits section of the evaluation: a call names a copy of this repository, a headwater binary or GitHub. It does not count every path outside the workspace. The corpus was not frozen between the two arms' sessions. docs/spec/12-check-layer.md changed after the recording, so each verdict is graded over a moved read set. The evaluation of that batch states every limit and the cost. This record stays open by the owner's ruling of 2026-09-30, until #1472 runs a discovery selection sized for power. The human half still has no instrument.

The campaign of 2026-10-03 ran the discovery selection at power, and it does not separate the arms. #1659 ran the three discovery probes at 118 repetitions each, in all six campaign arms. The present arm satisfied 137 of 354 sessions (38.7%) and the absent arm 152 of 354 (42.9%). The difference is +4.2 points for absent, in a 95% Newcombe interval of -3.0 to +11.4 points. The interval contains zero, so the batch does not show that the governance layer helps a cold session find a document. The reading has a limit: the recorder counted 177 paths outside the workspace in batch B. Each is a try and not a read, and the clause of #1659 that asks for 0 is not met. The evaluation states every figure. The human half still has no instrument, so this record stays open and it waits on a ruling. The earlier reason, the owner's ruling to wait for #1472, is spent, because the campaign ran.