SPINE_SCHEMA_v2.md

systems/SPINE_SCHEMA_v2.md

Spine entry schema v2 — the anchor-hub proposal (downstream-pipeline-derived)

Source: the spine-anchor-requirements analysis (wf_26795550-8fb) — 5 downstream consumers (book->region-pages, quest-writing, asset-gen, 3D-world, CVD-alignment) specified their needs; this is the synthesized updated spine-entry schema. Lean-via-anchoring (author narrative+alignment+synthesis in-spine; anchor bulk substrate into tiers/registries). Prose ~800-1200. Apply to the /spine-chapter skill + CHAPTER_BUILD_TARGET, then re-enrich Ch 2-13 to this bar.

---

I have everything I need. Here is the synthesis proposal.

---

Spine-Entry Schema v2.5 — Proposal: satisfying all five downstream consumers, lean-via-anchoring

Scope. This proposes STRUCTURE + anchoring only. Chapter/Component/numerical/hard-line/naming/CVD-thesis canon stays Josh's (cardinal rule). The current CH_11/CH_12 13-block structure is strong and is preserved verbatim; this adds 7 anchor-dense blocks and extends 2 existing ones, all consolidating substrate that today is scattered through prose. It aligns with the civilizational_marks field already added to T1_Build_Pipeline_Contracts §9.5 today (2026-07-08) and with the §9.6 scatter map, which already specs most of the target columns (region_chapter_set_array, scene_anchor entries, quest_anchor entries, boss_encounters_canonical_array, theme/voice recurrence_anchors) — so the new blocks are the AUTHORING SOURCE for scatters the schema already anticipated, not new schema.

Consumer shorthand: RP = Region Page / Book Processing (BPC §7/§8) · QW = Quest Writing (§6) · AG = Asset Generation (Meshy §2 / AIVA §3 / ElevenLabs §4 / SFX §10 / Cinematic §5) · 3D = UE5 World Build (§1) · XP = every pipeline (spine+CVD alignment frame).

---

1. The full ordered block list (v2.5)

A builder reads top-to-bottom. E = existing (keep), E+ = existing, extended, N = new.

Anti-duplication rule inside the entry: a value is stated ONCE. Each FK lives in Mechanical anchors (#18); the domain blocks (#6 Civilizational marks, #8 Key-sites, #17 Asset anchors) carry the connective JUDGMENT and reference back to the FK line rather than re-stating it. scripts + inscription_spine stay as FK lines in Mechanical anchors (their existing home); Asset anchors adds only the carrier-medium + props judgment on top — additive, low migration cost.

---

2. Prose scope and target anchor density

• Body prose (#4): hold BPC's 200–400w personal-arc discipline. It is NOT where value is added by growth.
• Authored blocks total (#3, 4, 6, 10, 11, 14, 16, 19, 20 + authored fragments): ~800–1200w as Josh guided — the growth from v2.4 is Main quest beats (~120–180w), Side-questlines (~80–140w), Civilizational marks authored L2 lines (~80–120w), Anti-fringe (~40–80w), the L3 expansion (~60–100w). Everything else is anchor-dense.
• Anchor blocks (#5, 8, 17, 18 + cross-links): target ≥1 resolvable [SRC:] per line; Mechanical anchors and Asset anchors approach 1 anchor per dashed line resolving to a registry ROW or a tier-doc heading.
• Whole-entry target density: ~1 resolvable [SRC:] per 25–40 words. Net word cost of v2.5 over v2.4 is ~250–350 authored words + ~14–18 new anchor lines — inside Josh's "800–1200 + mechanical blocks + anchors" envelope for an always-loaded hub.

---

3. Lean/quality balance rules — authored in-spine vs anchor-only

AUTHOR IN-SPINE (spine-native syntheses with no single registry home; the load-bearing framing that constrains every pipeline):

• Body prose (personal arc); Continuity anchor + End-state carry-forward (inter-chapter invariants).
• Main quest beats (dramatic spine + branch points) and Side-questlines per-chapter stage/gate/hook/reward (Questline_Index empty → in-spine is the only source).
• Three-layer separation contract; Protected-combat framing; the Anti-fringe one-line canonical framings (WHICH claims + the statement).
• Civilizational-marks L2 achievement lines (authored with [SRC:]).
• Reveal-discipline statements + awareness enums; reframed constructed names + pending-NAMING_METHOD flags.
• The connective judgment on every anchor line — which theme recurs, which creature is combat-permitted vs sacred-inviolate, which site hosts the boss, which entity is ethnographic vs stylized, which link fires forward.

ANCHOR ONLY — point, never inline (bulk substrate that lives in tiers/registries):

• Full citations / bibliography → T0_Source_Material_Registry + BOOK_ACQUISITION (that is region-page Section 16).
• Reframe RATIONALE → CVD §19.1/§19.2. Era-modification detail, coordinates, biome PCG enums, zone polygons → WGM / T0_Region_Index (Phase-6).
• Boss/NPC full persona + voice → BOSS_PERSONAS + T0_Character_Index + T0_Antagonist_Network_Registry. Loot tables → T0_RNG_Drop_Table + T0_Quest_Seed_Registry. Trade twelve-tier gates → T1_Trades_Master. Thread canon → T0_Thread_Grid + T1_Threads_22. Per-slot familiar mechanics → T0_Familiar_Registry. Weapon materials/stats → T0_Weapon_Registry. Creature morphology + architecture material specs → T0_Creature_Roster / WGM §14 / RCM.
• Owner-consistency fix (from the 3D consumer): all geography/site/biome anchors point to T1_World_Geography_Master, the geography authority — NOT RCM (CH_11 currently mis-anchors Exploration to RCM; CH_12 correctly points to WGM). RCM is substrate lore, not the geography owner.
• Empty-registry convention: where a target registry is a stub, the spine line carries [POPULATE→registry] + the RCM primary-source corpus pointer — enough for a build-time deep-research pass to synthesize faithful substrate. This is why the spine is currently the only per-chapter anchor ahead of asset build.

---

4. Gaps — what must exist for the spine to anchor into (cross-ref docs/FOUNDATION_ENHANCEMENT_ROADMAP.md)

No NEW registry is required by the schema change itself — civilizational_marks and every new block live in prose L2 + Mechanical/Asset anchors and introduce no new T0 column at this revision, so the fidelity gate is unaffected. The gaps are (a) already-specced Phase-6 COLUMNS not yet instantiated, (b) specced registries that are EMPTY (population), (c) proposed-but-uncreated owning tier docs.

(a) Columns already specced (BPC §1.6 / §7.6 / §9.6) — instantiate as empty CSV columns, populate at Phase-6:

• T0_Region_Index: biome_classification, canonical_zones, region_chapter_set_array — feed Entry-header biome line, Key-sites, region grouping. (BPC §7 region_id-format decision is still open in PIPELINE_ADAPTER.md.)
• T0_Chapter_Index: environmental_params, time-fluctuation window fields — feed Key-sites environmental_params.
• T0_Source_Material_Registry: region_tags, primary_category, canonical_retraction_status — feed substrate_sources and Anti-fringe blocks. (5 placeholder rows today; §8.5-conformant 28-col schema exists.)
• T0_Scene_Spec_Registry: per-chapter scene_anchor rows — feed Dialogue/scene anchors (registry ACTIVE, 16 rows, but no per-chapter rows yet).

(b) Specced-but-empty registries — POPULATION, tracked in the roadmap:

• T0_Questline_Index (DRAFT, empty) — the Side-questlines block is the only authorable SQ source until this lands (ROADMAP Grid 3 / RR every-10 audit).
• T0_Creature_Roster (0 rows), T0_Weapon_Registry (0 rows), T0_Familiar_Registry (DRAFT) — Asset anchors' creatures_and_fauna / weapons / familiars ride [POPULATE→] meanwhile. ROADMAP P0 (weapons: seat region-authentic arms — urumi/katar/keris/macuahuitl/khopesh/iklwa/kpinga), P1 (creature malevolent-being boss vein; familiar-combat spec).
• T0_Theme_Registry (12 rows, conforms) + T0_Voice_Registry (conforms) + T0_SFX_Registry (needs the §10.5 reconcile per PIPELINE_ADAPTER) — music_mood / voice_and_crowd / sfx_substrate anchors.
• T0_School_Registry (0 rows) — ROADMAP P0: create T1_Mystery_Schools_Master as the owning tier doc and populate the ~13 schools (CVD line 189); Civilizational-marks + Side-questlines (SQ8/SQ12) + L3 anchor into it.
• T0_Thread_Registry / T0_Thread_Grid (P1: registry empty; grid ~37% with copy-paste cross-bleed) — Threads-advancing + civilizational-marks Thread-13/17 co-anchors depend on a clean grid.

(c) Proposed-but-uncreated owning tier docs the v2.5 blocks want to point at (all in the roadmap):

• docs/ESOTERIC_GROUNDING.md (ROADMAP P0, paralleling SCIENCE_GROUNDING) — the L3 grounding law the expanded Three-layer block + Anti-fringe block anchor into; operationalizes the RCM's already-listed Faivre/Hanegraaff/Scholem/Burkert.
• Regional Combat Modifier table (Pillar 10 named artifact, ROADMAP P1) — Key-sites / boss / Asset weapon lines.
• A structured materia-medica / toxica field (ROADMAP P0/P1) — Civilizational-marks (medicine) + trade side-quest anchors.
• Thesis-discipline / Frame-A note (ROADMAP P0, CVD §2/§3 Components 1&3; the Nazca "landing diagram" strike at NW_0048; the Component-6 "energy vortex/orgone" purge; the "vril" §19-etymology note) — the canonical framings the Anti-fringe block cites. These are Josh-flagged CVD-wording calls; the spine block only NAMES the at-risk claim and points to the CVD entry, it does not author the reframe.

---

5. Migration note — evolving CH_02–13, the /spine-chapter skill, and CHAPTER_BUILD_TARGET.md

Principle: additive and non-destructive. The existing 13 blocks are strong and stay verbatim (the Protected-combat block, NPC roster + awareness, cross-cultural links, reveal-discipline, boss structure — the consumers flagged these as strengths to protect). v2.5 ADDS blocks; it does not re-derive entries.

CH_02–13 entries:

1. Retrofit CH_11 and CH_12 first to the v2.5 order — they are the mature reference entries; establishing the v2.5 exemplar on them gives the cascade and the critic a concrete target (the way CH_03/CH_42/CH_43 anchor the current bar).

2. Backfill CH_02–10 and CH_13 at their next Range-Review touch as a v2.5 uplift pass (RR-12/RR-22 have passed — treat as uplift, not re-authoring). Note civilizational_marks was mandated only today, so every CH_02–13 entry needs that block added regardless.

3. For each new block, existing entries harvest content that is ALREADY in their prose (the consumers confirmed the substrate is present but scattered): promote the Exploration bullet → Key-sites; lift boss phases + integrity_divergence + Body-prose fork → Main quest beats; lift the side_questlines half-lines → Side-questlines; lift scholars from Lore/Folklore/Protected prose → substrate_sources; lift the Dogon-Sirius/Griaule notes buried in Decisions → the standing Anti-fringe block; lift Thread-20 props + fauna/architecture from woven/boss prose → Asset anchors. Empty-registry lines carry [POPULATE→].

.claude/skills/spine-chapter/SKILL.md:

• Step 3 "the entry carries, in order" list → replace with the 20-block v2.5 order; add the 7 new blocks with their [SRC:] targets and the empty-registry [POPULATE→] convention.
• Extend the Mechanical-anchors field enumeration with substrate_sources, rewards, and the encounter_mode/zone note; add the owner-fix rule (geography → WGM, never RCM).
• Add the standing-block rule: Anti-fringe and Civilizational-marks are authored EVERY chapter, even when the answer is "none at-risk" / thin — so a builder never silently skips them.
• Extend the post-draft 9-item checklist to a ~14-item list covering the new blocks; update the normalize-doc critic rubric to check them (combat_permission tags present; ethnographic-vs-stylized flagged; substrate_sources resolves to BOOK_ACQUISITION/Source_Material; Anti-fringe present; per-path branch beats named; every asset line anchored or [POPULATE→]).
• Step 5 scatter: the new blocks feed §9.6 targets already specced (region_chapter_set_array, scene_anchor, quest_anchor, boss_encounters_canonical_array, theme/voice recurrence_anchors) — so no new schema; note that any scatter to a NEW Phase-6 column is a deliberate --emit-baseline re-emit in the same commit.

docs/proposals/CHAPTER_BUILD_TARGET.md:

• §5 civilizational_marks already exists — keep. ADD sections for: the Asset-anchor manifest bar (one anchored line per domain; the CH_11/CH_12 two-python permitted-vs-sacred contrast is the model combat_permission exemplar), the Main-quest-beats + Side-questlines authored bar (ordered spine + named branch points; per-SQ stage/gate/hook/reward), the Anti-fringe standing block (the one-line canonical framing method), and the Key-sites spatial bar (site→gameplay_function→L1/L2/L3→zone-seed). Point each to the worked CH_11/CH_12 lines as the reference build.

T1_Build_Pipeline_Contracts §9.5/§9.6: no schema edit required — like civilizational_marks, the new blocks are prose/anchor additions that introduce no new T0 column at this revision, so the fidelity gate is unaffected; the §9.6 scatter map already carries the target columns. A future contract revision can promote the recurring lines (e.g. substrate_sources, environmental_params) to explicit schema fields once Phase-6 operational friction confirms the shape — the contract's own minimum-viable-defer clause.

---

Bottom line. Five consumers, three real structural gaps, one design move: consolidate the already-present-but-scattered substrate into 7 anchor-dense blocks (Civilizational marks, Key-sites, Main-quest-beats, Side-questlines, Dialogue anchors, Anti-fringe flags, Asset anchors) + 2 header/mechanical extensions, authoring in-spine ONLY the load-bearing framing (the dramatic spine, the protected-combat and Frame-A disciplines, the per-domain connective judgment) and POINTING everything else into the owning tier docs and registries. The spine stays lean enough to be the always-loaded hub because leanness comes from anchoring, never from thinning — and the schema already anticipated most of the scatter targets, so the change is population and authoring discipline, not new canon.

Key files read: docs/spine/CH_11.md, docs/spine/CH_12.md, docs/PIPELINE_ADAPTER.md, _source/01_Tier_1_Foundation/T1_Build_Pipeline_Contracts [ACTIVE v1.0].md §9.5/§9.6, docs/proposals/CHAPTER_BUILD_TARGET.md, docs/FOUNDATION_ENHANCEMENT_ROADMAP.md, .claude/skills/spine-chapter/SKILL.md, and the live registries/ roster.

Generated by harness/site/structure_site.py — the URL path is the repo path. review root