[EPIC] Procedural Level Pipeline: lore → structure → greybox → registered plate → derived manifests

[100yenadmin]

Owner-ratified direction (2026-07-11): generating FUNCTIONAL LEVELS is the foundation priority — level design first, art derives from it. This epic ties together the pieces that now exist into one repeatable pipeline, honoring the engine-as-sole-writer invariant throughout.

The pipeline (each stage's output is the next stage's input)

1. STRUCTURE — engine-side scene_grid generators (lore-aware, per location kind) OR authoring-time Unity generators (DunGen room-graphs / Tessera Pro WFC — both owner-purchased 2026-07-11) whose layout data is EXPORTED to json. Unity generators are authoring accelerators only; the engine fixture remains the single written truth.
2. GREYBOX — layout json → 3D proxy geometry (shape-appropriate proxies: boxes for masonry, cylinders/cones for organics — box-trees read as buildings to depth models, measured PR Plate sprint: CB-FOREST evidence — ARM C LoRA does NOT lift the outdoor 6.0 cap #1495) → greybox depth/normal renders + geometry json.
3. PLATE — the codified registered pipeline (docs/roadmap/PLATE-RECIPE-DECISION.md): flux depth-CN base → style pass → registration gate → 5-scorer panel → promote.py gate.
4. MANIFESTS — tools/derive_room_manifest.py (in flight): geometry → footprint cells + occlusion regions + walkable set → engine fixture + player occluders + coherence gate. One source of truth; paint/collision/occlusion cannot drift.
5. RUNTIME — plates_manifest.json entry + plate PNG (the W5E registry): add a room = add data, no scene work.

Stage status

• 1 (engine half): EXISTS (scene_grid per-kind generators). 1 (Unity half): SPIKE — DunGen export contract (owner hunch, prioritized), Tessera Pro follow-up.
• 2: half-exists (hand-built per room today; standardize the exporter).
• 3: DONE (plate sprint).
• 4: IN FLIGHT (eval-upgrade lane).
• 5: DONE (walkable slice).

Invariants

Engine = sole writer of grid truth (generators export TO fixtures, never run at runtime). Every generated room passes: registration gate, grid-paint coherence gate, journey eval. Panel + promote gate before adoption.

Definition of done (epic)

A new room goes from "generate layout" to "walkable in the player with correct collision/occlusion and a paneled registered plate" via commands only, no hand calibration, in under an hour of wall-clock.

[evaos-code-review-bot]

evaOS issue enrichment

Issue: #1508 - [EPIC] Procedural Level Pipeline: lore → structure → greybox → registered plate → derived manifests
State: open

Related issues/PRs: #1495.
Existing labels: none.
Suggested labels: security, docs, tests.
Suggested owners: none.

Related context

• Plate sprint: CB-FOREST evidence — ARM C LoRA does NOT lift the outdoor 6.0 cap #1495 - mentioned in issue metadata; inspect for dependency, duplicate, or prior-decision risk.

Agent-start packet

Problem shape: [EPIC] Procedural Level Pipeline: lore → structure → greybox → registered plate → derived manifests - Owner-ratified direction (2026-07-11): generating FUNCTIONAL LEVELS is the foundation priority — level design first, art derives from it. This epic ties together the pieces that now exist into one repeatable pipeline,
Product fit: Product/UX-sensitive issue; review against repo vision and user flow quality, not only technical P0/P1 risk.
Implementation wedge: Ship the smallest user-visible path or prototype that can be smoke-tested against the intended experience.

Build / borrow / buy scan:

• Build: identify the smallest implementation wedge that proves the issue without broad rollout.
• Borrow: inspect same-repo prior art, repo memory, GitNexus context, and allowlisted cross-repo examples first.
• Buy/use: run capped external OSS/library/API/current-market research with citations, observed dates, and why each source matters.

Candidate sources:

• VISION.md / repo policy / repo-memory.md / GitNexus packet.
• Same-repo issues and PRs referenced by issue metadata.
• Allowlisted cross-repo GitHub examples with direct issue/PR/repo URLs.
• External OSS repositories or library/API docs with freshness dates.
• Current market/product examples, capped and cited, only when they change implementation direction.

Acceptance criteria:

• Add explicit acceptance criteria before implementation is considered done.
• Include source citations for any external OSS/library/current-market claims.
• Keep issue-enrichment comments sticky and idempotent.
• Keep labels, owners, reviewers, and roadmap fields suggestion-only.

Proof plan:

• Focused unit/fixture test for the changed planner behavior.
• Dry-run issue-enrichment evidence packet before live comment posting.
• Redaction check for issue body, citations, and generated markdown.
• UX/product smoke or screenshot proof when implementation changes user flow.
• Contract or degraded-mode proof before runtime promotion.

Known traps:

• Do not treat related links as proof by themselves; each source needs a reason why it matters.
• Do not scan old issue backlogs or exceed repo-level enrichment throttles.
• External research must stay capped, cited, fresh-dated, and redacted.
• Technical severity alone can miss product/UX regressions; include product-manager review framing.

Non-goals:

• Do not auto-apply labels, owners, reviewers, roadmap fields, or milestones.
• Do not bulk-enrich old backlog issues from this planner output.
• Do not claim external market or OSS research was performed unless cited sources are present.

Context-source taxonomy:

• [enabled] vision_repo_policy_memory_gitnexus: Use repo vision, policy, memory, and GitNexus first before external research.
• [enabled] same_repo_issues_prs: Issue metadata contains same-repo references.
• [enabled] allowlisted_cross_repo_github: Triggered by issue class; keep to configured allowlist and source caps.
• [enabled] external_oss_examples: Use for build/borrow/buy precedent with citations and observed dates.
• [enabled] library_api_docs: Use when implementation may depend on a module, SDK, API, or framework choice.
• [enabled] current_market_examples: Use for product/UX/market positioning issues with freshness dates.
• [enabled] internal_mcp_docs: Use internal docs/MCP context when available; missing context is degraded, not blocking.

Validation suggestions

• Confirm owner, acceptance criteria, and validation evidence before implementation.

Triage gaps

• Acceptance criteria or validation evidence not detected in issue body.

No labels, owners, reviewers, or roadmap fields were changed by this bot.

[100yenadmin]

LARGE SPACES stage ruled (PR #1527 memo): rooms atomic at native density; towns = generator-graph layouts with door-cross + masked boundaries; hybrid seam recipe = special-case vistas. NEW PRINCIPLE from CRYPT-REPLICATE (PR #1528, honest 7.1 ceiling vs 7.8 incumbent): paint richness follows GEOMETRY richness — the honest path past the drift-rich incumbents is authoring denser greyboxes (more prop volumes = richer paint AND richer collision), not more cosmetic style passes. Queued: richer crypt greybox iteration (6-10 additional prop volumes) through the standard loop.