011: measure the graph against every facet a module carries #11

Merged
jschoubben merged 1 commits from research/011-facet-coverage into main 2026-09-05 00:53:20 +00:00
Owner

Adds a coverage section to research 011: the five declarations describe how modules relate, but a module is more than its relations — so the as-is manifest surface becomes the checklist the effort is measured against, facet by facet.

Each candidate gap is weighed rather than adopted:

  • Covered: resources (exists), host-applied facets (ADR 0043 typed resources — closing the issue-003 class), environment, migrations (ADR 0006 stands; the two-kinds distinction must survive), scheduled tasks.
  • Probably dissolves: per-node selections (into requires a node capability + configuration), cross-component contributions (into declared resources, once host absorption is accounted for) — both to confirm, not assume.
  • Genuinely open, added to the open-questions table: where the tool surface lives (the largest unscoped facet — tier 3's "no logic" rule vs module-specific tools), when a provides-edge is true (issue 007; mandatory verifiers risk trivial ones), and whether module-to-module contributions need an edge at all.

https://claude.ai/code/session_01LrgweAeERJYBg88c5cKDzF

Adds a coverage section to research 011: the five declarations describe how modules relate, but a module is more than its relations — so the as-is manifest surface becomes the checklist the effort is measured against, facet by facet. Each candidate gap is weighed rather than adopted: - **Covered**: resources (exists), host-applied facets (ADR 0043 typed resources — closing the issue-003 class), environment, migrations (ADR 0006 stands; the two-kinds distinction must survive), scheduled tasks. - **Probably dissolves**: per-node selections (into `requires a node capability` + configuration), cross-component contributions (into declared resources, once host absorption is accounted for) — both to confirm, not assume. - **Genuinely open**, added to the open-questions table: where the tool surface lives (the largest unscoped facet — tier 3's "no logic" rule vs module-specific tools), when a provides-edge is true (issue 007; mandatory verifiers risk trivial ones), and whether module-to-module contributions need an edge at all. https://claude.ai/code/session_01LrgweAeERJYBg88c5cKDzF
jschoubben merged commit 1a83ed28fa into main 2026-09-05 00:53:20 +00:00
jschoubben deleted branch research/011-facet-coverage 2026-09-05 00:53:20 +00:00
Sign in to join this conversation.
No Reviewers
No labels
1 Participants
Notifications
Due Date
No due date set.
Dependencies

No dependencies set.

Reference: novox/hq#11