WORK-617
ID:WORK-617Status:cancelled

character, realm and faction drop their transforms for the slot declaration

The second proving set for SPEC-143: the storytelling entity runes whose sections arrive as already-rendered child runes (emitTag). Together with WORK-616 this covers both sides of the emitTag split, which is what proves the mechanism rather than one shape of it.

Same gate (D7): no drift in contracts or the SEO baseline, per rune.

recipe is the standing proof that the escape hatch is still needed (D3). It is not migrated, and it is not bent to fit.

Priority:mediumComplexity:moderateMilestone:v0.39.0Source:SPEC-143
main View source

Criteria completion

Criteria completion: 0 of 6 (0%) checked; tracking started on Oct 7, no incremental history yet0%25%50%75%100%Oct 7Oct 11

Tracking started Oct 7 — check back for trends.

Relationships

Implements 1
Related 8
Branches 3
main current cancelled
claude/post-v0.38-roadmap-jb7pie cancelledchangeset-release/main readyclaude/v039-slot-labelling cancelled
History 4
  1. cfe714f
    statusready→cancelled
    by bjornolofandersson
  2. 5191d3e
    Content editedby Claude
    docs(plan): cancel WORK-617, defer storytelling to composed runes
  3. 50daca8
    Created (ready)by bjornolofandersson
  4. 1046f41
    Content editedby Claude
    docs(plan): add v0.39.0 milestone, Declarations part two

Sequencing note

WORK-612 rewrites the same plugin's registration. The two are independent; land them sequentially rather than on parallel branches.

Acceptance Criteria

  • character, realm and faction have no transform for the entity rune, and refrakt contracts --check plus npm run seo:baseline:check report no drift for any of them
  • buildStoryContent is deleted, not merely unused
  • recipe still has its transform and is not contorted to fit (D3)
  • The cross-page pipeline's registry, extractTitle and breadcrumb reads produce identical results for storytelling content
  • Each migrated declaration round-trips through JSON.parse(JSON.stringify(…)) unchanged
  • The total transform() line count across the tag files is recorded after the migration, beside WORK-616's before figure

Blocked by

  • WORK-614

References

  • SPEC-143 — D3, D5, D7

Resolution

Completed: 2026-10-07

Cancelled, not done or superseded, by the repository owner's decision on 2026-10-07.

Why

character, realm and faction fail SPEC-143's D4 family test ("does every output slot come from exactly one resolved field, wrapped but not restructured?"). This was checked against their real output on claude/v039-slot-labelling (refrakt-md/refrakt#668):

  • realm / faction emit <div data-name="scene"><img data-name="sceneImage"></div>. That is a named node inside a named node, which D2 forbids, and the schema.org image row reads sceneImage, so it cannot be flattened without moving the SEO baseline. extractScene also splits the one scene field into either the scene image or extra body prose by node kind (the pick capability the spec declined).
  • character filters its rendered portrait with .tag('img').limit(1), so a ![x](placeholder:portrait) image is dropped today. A plain region would emit it.
  • All three filter rendered children to the section rune type, and they merge non-section preamble runes into body only when there are no sections (and drop them when there are). That is a conditional across two slots. They also set data-field="section" on the rendered child runes.

Reproducing them byte for byte would need nesting, pick, filter and a cross-slot conditional, all of which SPEC-143 rules out (D2, D4, "Capabilities considered, and declined").

Where it goes

The composed-runes work (SPEC-145 / SPEC-147), which replaces these runes rather than declaring their current transforms. buildStoryContent and the storytelling transforms stay until then.

What SPEC-143 still has

  • The mechanism is proven on the resolved-entry (non-emitTag) side by WORK-616 alone.
  • The emitTag side of the D5 split is covered by the mechanism's own tests (packages/runes/test/slots.test.ts), not by a migrated production rune.
  • The transform() line-count after figure is recorded in WORK-616 instead: 129 transforms / 6,385 lines before, 126 / 6,282 after.

Notes

  • The child sub-runes (character-section, realm-section, faction-section) looked like they would fit: an attribute value in a span, plus a body region. Not migrated, not filed.