Imported from organvm-iv-taxis/a-i--skills (distributions/claude/skills/domain-ideal-whole-substrate/SKILL.md). Install upstream with npx skills add organvm-iv-taxis/a-i--skills --skill domain-ideal-whole-substrate. Copyright stays with the author (MIT).

Domain Ideal-Whole Substrate (DIWS)

The substrate beneath every flag-pierce. Where Product Domain Engine (PDE) handles the product layer (logos / ethos / pathos / kairos), this skill handles the domain layer beneath. Every domain — chess, fitness, voodoo, education, design — composes from the same eight strata under the same four operators, with load distribution shifting by domain.

User's foundational question: "What elements, systems, structures, environments, agents so forth make the ideal whole (be it fitness, chess, or voodoo)?"

DIWS answers it as substrate, not as workflow.

What This Is

A substrate, not a workflow. PDE is a 5-phase formalization protocol that consumes a domain and produces a product. DIWS is what the domain itself is before any product gets built on top of it. PDE pulls from DIWS substrate; DIWS does not pull from PDE.

DIWS does two things PDE cannot:

1. Cross-portfolio reuse — Phase 0 audits ~/Workspace/ for existing engines (3-tier taxonomy: domain / meta / consultant) before any new domain instantiates. Engine + skin pattern, not from-scratch every time.
2. Stretching rack — Phase 0.5 runs the gap-map operator across N concurrent DIWS instances simultaneously. Surfaces holes (under-developed dimensions) and fat (redundant scaffolding) at portfolio scale, not just per-domain.

You invoke this skill when a flag has been pierced or is about to be — a friend's domain, a research interest, a consulting engagement, a side practice. The goal is to load the substrate ahead of any product-building work so that PDE / synthesis / cross-pollination diagnoses operate over filled scaffolding rather than empty types.

Invariants

These hold across every flag-pierce, regardless of domain:

1. Phase 0 portfolio-audit runs BEFORE layer-1 ontology — never scratch-build what an existing engine can re-skin (rule 41: audit before building)
2. Phase 0.5 engine-overlap board comes between Phase 0 and layer-1 — visualize candidate engines + gaps before scaffolding (rule 36: seed not specification)
3. All 8 strata instantiate — even if a stratum is sparse for the domain (load shifts, layers don't disappear)
4. All 4 operators fire on every instantiation — selfish-altruistic + magnetic + portfolio + reflexive (Tenet Protocol — both directions simultaneously)
5. Layer 4 (gap-map) and Layer 8 (contribution) close as a loop — heist + contribute round-trip; never steal without giving back
6. Layer 8 produces dissertation-grade research output — academic capital is substrate, not garnish
7. Reflexive Operator produces domain-meta-study.md alongside the 8 stratum artifacts — every build co-arises with a study-of-build
8. Capital flow is N-channel, not 1-channel — monetary is one of many; meta-skill capital, audience cross-flow, tool extraction, work-improvement all count

The 8 Strata

Every domain composes from these. See references/8-strata-spec.md for full per-stratum spec.

Voodoo loads layer 8 heavily (ethical floor, anti-extraction guard). Chess and fitness share a similar load profile. Both shapes validate the schema — generative across radically dissimilar domains rather than overfit to one.

The 4 Operators

See references/4-operators-spec.md for full spec. All four fire simultaneously on every flag-pierce.

1. Selfish-altruistic loop (single-engagement)

Learn-for-self → Quality-of-product → Service-to-friends →
Friends-elevated → Network-effect → Self-elevated → More-learning

The selfishness is a feature. It locks the user into being a real practitioner rather than a curator. Without the lock, the system curates dead.

2. Magnetic membrane (single-domain, two-way)

Flag pierces →
  INTERNAL: magnetize all applicable, transmute through user's lens, alchemize toward ideal
  EXTERNAL: contribute to community

Pull-in = refinery. Push-out = gift. Per Tenet Protocol — both directions simultaneously, never one without the other.

3. Portfolio Operator (cross-portfolio, two-way) — v2 + v2.1

Push (v2): every flag-pierce produces three cross-flows simultaneously:

• meta-skill extraction (writing / voice / community-building)
• domain-pair overlaps (each new domain creates N-1 publishable conversation surfaces)
• tool/audience cross-flow (engines amortize across portfolio nodes)

Pull (v2.1): every flag-pierce begins with portfolio-audit — scan ~/Workspace/ for engines / skills / primitives that match candidate-domain shape; only enter layers 1-8 after portfolio-reuse-map.md is filled.

4. Reflexive Operator (build + study, parallel) — v2.2

Every domain instantiation produces domain-meta-study.md alongside the 8 stratum artifacts. Where the other three operators govern flow, this one governs recursion — the thing and the thing-studied co-arise. Per Tenet Protocol, every build-action triggers its opposing reflection-action simultaneously.

The Use Protocol

Sequential within a single instantiation; recursive across portfolio lifecycle.

Phase 0 — Portfolio audit (PULL)

Run scripts/audit-portfolio.sh. Globs ~/Workspace/ for existing engines and classifies them into the 3-tier engine taxonomy (see references/engine-taxonomy.md):

• Domain engines — re-skinnable within similar domain (BODI 4-level funnel, spiral renderer, landing-engine)
• Meta-engines — cross-domain pattern (cross-pollination diagnosis, 75-person Constellation, PDE skill, Bridge Content templates, Discord rituals)
• Consultant engines — every engagement regardless of domain (knowledge base, application pipeline, plan-mode discipline, IRF/MEMORY/chezmoi protocols, conversation-corpus pipeline)

Output: portfolio-reuse-map.md showing which engines re-skin, which need invention, which are skin-only domain-specific work.

Phase 0.5 — Engine overlap board (THE STRETCHING RACK)

Run scripts/portfolio-gap-audit.sh. Operates across N concurrent DIWS instances simultaneously to surface holes and fat at portfolio scale.

For each axis (the 8 strata + 4 operators), the script reports:

• Holes — under-developed in 1+ instances; transplantable from a sibling instance
• Fat — over-engineered in 1+ instances; consolidatable into a shared engine
• N-way overlap — concept present in ≥2 instances; promotion candidate to meta-engine

Output: holes-fat-report.md. Refines portfolio reuse from listing engines to seeing them in spatial relation, identifying white-space across the whole portfolio.

Phase 1-8 — Stratum instantiation

Walk each stratum in order, filling its template from assets/ into the domain's instantiation directory. Use references/8-strata-spec.md for canonical questions.

Order is canonical but not strictly serial — strata 1-3 can be parallelized; 4 depends on 3; 5-8 are largely independent once 1-4 are in place.

Phase 9 — Operators fire (REFLEXIVE)

After stratum artifacts ship:

• Selfish-altruistic loop: confirm the user is in practitioner-mode, not curator-mode (gate)
• Magnetic membrane: confirm both INTERNAL and EXTERNAL outputs exist (no one-way pierces)
• Portfolio Operator: write portfolio-resonance.md — 3 cross-flows fired
• Reflexive Operator: write domain-meta-study.md alongside the 8 strata

Phase 10 — Cross-pollination diagnosis (synthesis)

Run scripts/portfolio-gap-audit.sh with this new instance + sibling instances. Surfaces N-1 cross-pollination opportunities by default. Each becomes a candidate essay, video, or joint-product opportunity.

Engine + Skin Pattern

Engine = portfolio-pull primitive (re-skinnable)
Skin   = domain-specific 8-stratum instantiation
New code = only what neither engine nor skin can express

Target ratio: next 5 friend-engagements consume 60–80% engine + 20–40% skin, not 100% from-scratch. Phase 0 enforces this.

Layer 4 ↔ Layer 8 Heist + Contribute Loop

Layer 4 names heist targets — homages worth remixing. Layer 8 closes the loop by pushing improvements back upstream to the same competitors. Heist becomes relationship; relationship becomes co-development. Never steal without giving back.

Intelligence-gathering is ongoing protocol, not one-shot. Continuous monitoring of gap-map targets means the contribution charter has continuous output.

Composition with PDE

DIWS sits beneath PDE. PDE invocations consume a filled DIWS instance:

DIWS Phase 0  →  audit portfolio
DIWS Phase 0.5 →  stretching-rack diagnosis
DIWS Phase 1-8 →  fill 8 strata for domain
DIWS Phase 9  →  fire 4 operators
─────────────────────────────────────────
PDE Phase 1   →  IDENTIFY (consumes DIWS Stratum 5 agent-fleet)
PDE Phase 2   →  MAP (consumes DIWS Strata 1-4)
PDE Phase 3   →  ENCODE (types, schemas, tests)
PDE Phase 4   →  EXPRESS (4 rhetorical modes)
PDE Phase 5   →  DEPLOY
─────────────────────────────────────────
DIWS Phase 10 →  cross-pollination diagnosis (post-PDE)

PDE Phase 1 (IDENTIFY) was previously rebuilding the agent census from scratch every time. With DIWS, Stratum 5 (domain-agent-fleet.yaml) is the canonical source — PDE consumes it.

Composition with MSP (Modular Synthesis Philosophy)

See references/diws-msp-mapping.md for the full isomorphism. Briefly:

MSP names how the patches compose. DIWS names what the modules contain.

Modes

Distinct invocation modes the skill supports:

• mode portfolio-pull — Phase 0 driver; produces portfolio-reuse-map.md
• mode portfolio-resonance — Phase 0.5 driver; runs stretching rack across N instances
• mode reflexive-study — produces domain-meta-study.md alongside any stratum artifact
• mode engine-extract — when an instance reveals a generalizable pattern, flag for promotion to meta-engine or consultant-engine (user-authorized)
• mode full-instantiation — all phases 0 → 10 for a single domain
• mode cross-pollination-diagnosis — focused Phase 10 run between 2+ existing instances

Proof Instances

Stress-tested fills under proof-instances/:

• chess/ — Hokage Chess (Rob Bonavoglia client) as study source
• wellness/ — BODI fitness / Maddie Elevate Align as study source
• education/ — Jessica stub (DIWS schema v2.2 reference)
• voodoo/ — outlier validator (loads layer 8 heavily; design-only)

The chess and wellness instances are READ-ONLY studies of existing client work — they extract the substrate already running underneath those repos without modifying source repos. The education stub demonstrates how a pre-active flag is loaded. The voodoo outlier proves the schema generates across radically dissimilar domains.

Out-of-Scope

• DIWS does NOT build products — that's PDE's job
• DIWS does NOT execute outreach to constellation members — that's the user's call
• DIWS does NOT modify source-repo state when building proof instances — substrate fills live in this skill's directory only
• DIWS does NOT auto-promote engines to higher tiers — the user authorizes promotion via mode engine-extract

Verification (after running on a new flag-pierce)

• portfolio-reuse-map.md exists; ≥3 reuse decisions documented
• holes-fat-report.md exists; ≥1 entry per axis per instance
• All 8 stratum artifacts exist in the instantiation directory
• domain-meta-study.md exists (Reflexive Operator output)
• portfolio-resonance.md exists; ≥3 cross-flows named
• Engine taxonomy declared (which tier each consumed engine sits in)
• Heist + contribute round-trip declared in Layer 4 ↔ 8

Rules of Engagement

• Dissertation-grade default: Layer 8 output is substrate, not garnish
• Heist + Contribute: stealing an idea requires an equal or greater upstream gift
• Engine Taxonomy: categorize reusables as Domain, Meta, or Consultant tier
• All 4 operators or none: Tenet Protocol — partial firing is system failure
• Read-only on source repos: when building proof instances from existing client work, never write into source repos

Cross-references

• Plan: ~/.claude/plans/2026-04-25-domain-ideal-whole-substrate-design.md (v2.2 spec source)
• Sibling skill: ~/Workspace/a-i--skills/skills/project-management/product-domain-engine/SKILL.md
• MSP: ~/Workspace/a-i--skills/skills/creative/modular-synthesis-philosophy/SKILL.md
• Stretching-rack genesis: ~/.claude/plans/okay-so-now-harmonic-kettle.md (Stream Σ definition)
• Engine taxonomy: references/engine-taxonomy.md
• 8 strata: references/8-strata-spec.md
• 4 operators: references/4-operators-spec.md
• Portfolio composition: references/portfolio-composition-map.md
• DIWS↔MSP isomorphism: references/diws-msp-mapping.md