aura run --params: verify it closes free knobs; make the scalar-cell shape self-explaining #340

New Issue

2026-07-24T20:09:28+02:00

claude commented

2026-07-24 20:09:28 +02:00

Observed (source-blind fieldtest, #317 cycle close, 2026-07-24)

Orthogonal, pre-existing, NOT introduced by #317 — recorded separately so it
is not conflated with the open-pattern behaviour.

The fieldtester (working from the public interface only) could not close a
blueprint's free knobs with aura run --params '[…]': the run still refused
with "3 free knob(s)", exit 2. The flag's help is terse ("Blueprint params
(JSON scalar-cell array; .json mode)") — it does not say what a scalar-cell
array looks like, in which order the cells apply, or whether the flag targets
the same knob set sweep --list-axes reports.

Ask

Verify whether the flag works as documented (fieldtester usage error vs real
defect), and either way make the surface self-explaining (C29): a worked
example in the help text or the authoring guide, and a refusal that names the
expected shape when the array does not parse/apply.

Evidence: fieldtests/composites-use/TRANSCRIPT.md (#317 cycle corpus).
context: routed from the composites-use fieldtest at #317 audit close.

## Observed (source-blind fieldtest, #317 cycle close, 2026-07-24) Orthogonal, pre-existing, NOT introduced by #317 — recorded separately so it is not conflated with the open-pattern behaviour. The fieldtester (working from the public interface only) could not close a blueprint's free knobs with `aura run --params '[…]'`: the run still refused with "3 free knob(s)", exit 2. The flag's help is terse ("Blueprint params (JSON scalar-cell array; .json mode)") — it does not say what a scalar-cell array looks like, in which order the cells apply, or whether the flag targets the same knob set `sweep --list-axes` reports. ## Ask Verify whether the flag works as documented (fieldtester usage error vs real defect), and either way make the surface self-explaining (C29): a worked example in the help text or the authoring guide, and a refusal that names the expected shape when the array does not parse/apply. Evidence: fieldtests/composites-use/TRANSCRIPT.md (#317 cycle corpus). context: routed from the composites-use fieldtest at #317 audit close.

claude referenced this issue

2026-07-24 20:10:02 +02:00

Make composites data-authorable — a named, registered subgraph referencable like a node #317

Brummel referenced this issue from a commit

2026-07-24 20:19:52 +02:00

fieldtest: composites-use — 4 examples, 1 bug / 1 friction / 3 spec-gap / 3 working

claude referenced this issue

2026-07-25 17:46:48 +02:00

Retire the research-verb sugar in favour of document-first authoring over a self-explaining CLI #319

Sign in to join this conversation.