WORK-621
ID:WORK-621Status:done

data-owner and data-slot survive a primitive's transform

SPEC-145 D10a, landed ahead of composition so it can be proved inert.

Markdoc keeps only the attributes a node's schema declares. A marker set on a node before a primitive transforms it is therefore dropped, unless every node and tag schema declares it. Declare data-owner and data-slot once, at config assembly, on every node and tag schema, never per rune. Teach the schema-table resolvers (packages/runes/src/lib/schema-table.ts) to admit a node past a boundary by data-owner plus data-slot. A primitive's own data-name on the same node is left alone.

In the output, releaseOwnedNodes strips data-owner and keeps data-slot (D10a step 4, D10c).

Nothing sets either attribute yet, so the gate is that nothing moves. The survival test runs the markers through every rune in D12's placement set. That set is "no peer schema, no requiresParent". Any rune that drops either attribute is a finding, named by the test.

Priority:highComplexity:moderateMilestone:v0.40.0Source:SPEC-145
claude/v040-slot-markers View source

Criteria completion

Criteria completion: 5 of 5 (100%) checked; history from Oct 8 to Oct 80%25%50%75%100%Oct 8Oct 8

Relationships

Implements 1
Related 3
Branches 2
claude/v040-slot-markers current done
claude/post-v0.38-roadmap-jb7pie donemain done
History 2
  1. b5a5fc7
    • ☑ `data-owner` and `data-slot` are declared once at config assembly on every node and tag schema, not per rune, and a placed node keeps both through the primitive's transform (D10a)
    • ☑ The resolvers admit a node past a boundary by `data-owner` plus `data-slot`, and a primitive's own `data-name` on the same node is left untouched (D10a)
    • ☑ A survival test runs a marked node through every rune in D12's placement set and fails naming any rune that drops either attribute (D10a)
    • ☑ `releaseOwnedNodes` strips `data-owner` and leaves `data-slot` in the rendered output
    • ☑ `npm run seo:baseline:check` and `refrakt contracts --check` report no drift on either contract copy — nothing sets the markers yet
    by bjornolofandersson
  2. bf59a0e
    Created (in-progress)by bjornolofandersson

Acceptance Criteria

  • data-owner and data-slot are declared once at config assembly on every node and tag schema, not per rune, and a placed node keeps both through the primitive's transform (D10a)
  • The resolvers admit a node past a boundary by data-owner plus data-slot, and a primitive's own data-name on the same node is left untouched (D10a)
  • A survival test runs a marked node through every rune in D12's placement set and fails naming any rune that drops either attribute (D10a)
  • releaseOwnedNodes strips data-owner and leaves data-slot in the rendered output
  • npm run seo:baseline:check and refrakt contracts --check report no drift on either contract copy — nothing sets the markers yet

Resolution

Completed: 2026-10-08

Branch: claude/v040-slot-markers PR: refrakt-md/refrakt#680

What was done

  • packages/runes/src/lib/slot-markers.ts — declareSlotMarkers / declareSlotMarkersOnNodes: copy every tag/node schema (Markdoc's fallback nodes included) with data-owner and data-slot declared; memoised per map; shared exports never mutated.
  • packages/content/src/site.ts — assembleMarkdocSchemas(): the one place core + plugin schemas are assembled, used by both the page transform config and the preprocess hooks' embedConfig.
  • packages/runes/src/lib/schema-side-table.ts — SchemaSideTable, which resolves an assembled copy to its origin; the seven schema-keyed side tables (schemaPreprocessors, schemaTables, schemaRuneStructures, schemaEmits, schemaRegisters, schemaBasePresets, registers-source cache) now use it.
  • packages/runes/src/lib/schema-table.ts — SLOT_ATTR; isSlotted admits a node by owner + slot in findAllByName and findChildren; isMine splits the namespaces for a node carrying both markers.
  • Tests: packages/runes/test/schema-table.test.ts (slot resolution, release keeps data-slot), packages/runes/test/slot-markers.test.ts (declaration, non-mutation, side-table fallback, guard against raw WeakMap<Schema), packages/content/test/slot-marker-survival.test.ts (D12 set derived from requiresParent + schema types; seven node kinds through 92 runes).
  • Changeset: minor for @refrakt-md/runes and @refrakt-md/content.

Findings

  • SPEC-146 isMine vs D10a. "Extends, not replaces" did not hold as written: isMine rejected a node marked for another rune even among the primitive's own nodes, so a primitive could not resolve its own data-name on a placed node. Measured: figure with an image marked for character lost contentUrl. Resolved narrowly: a node carrying both markers keeps its data-name / data-field / data-rune in the primitive's namespace, and the owner reaches it only by data-slot. A node with data-owner alone keeps SPEC-146's behaviour. Inert today; confirm before WORK-622.
  • Schema identity. Copying schemas at assembly broke every identity-keyed side table ({% include %} reached the transform unresolved). Fixed with SchemaSideTable.
  • Drops (kept out of D12's set for that node kind, per D10a step 5, listed in NOT_YET_PLACEABLE): nav list, tabs heading/list, budget-category list, conversation blockquote, reveal heading, form paragraph/list/heading/blockquote, comparison heading/list, symbol-group list, changelog heading, preview fence. Each rebuilds that kind of node into its own structure.
  • snippet, data and include are resolved in preprocess and cannot be probed by a transform; carrying a marker through them is WORK-622's concern.
  • D9's "peer" is relational; the derivation treats every non-media type as a peer, which is conservative.

Notes

  • SEO baseline and both contract copies unchanged; npm test 5342 passed.