CoChalet: Fifteen-Layer Source-Body and Authority Master Analysis V2.1

Original evidence date: 2026-07-12
Legal-authority update: 2026-07-13
Verdict: REVISE
State: STAGED / VERIFIED-TO-MECHANICAL-SCOPE / NOT PROMOTED
Evidence ceiling: COMPLETE_INDEX_AND_SOURCE-BODY_POPULATION / TCJ_COUNSEL_RECORD_INTEGRATED / CLEARANCE_GATES_OPEN
Audience: Founder and named functional owners
Model disclosure: The requested Sol 5.6 Ultra identity was not independently verifiable and is not claimed.

Executive Summary

• V2.1 adds the audited July 13 counsel record to the authority spine without rewriting the July 12 corpus baseline. The 67-page response contains 169 verified highlights: 136 green counsel-response passages and 33 yellow passages. The yellow set is 32 substantive disposition items plus one formatting marker. This record materially improves legal traceability, but it does not turn the qualified opinion into unconditional clearance.
• CoChalet is best understood as a Quebec-native, deeded vacation-home co-ownership system with a hospitality and operating layer, not merely a real-estate marketplace. The corpus repeatedly describes a product that combines registered property ownership, scheduled personal use, professional operations, owner services, and digital coordination. The intended customer is an owner-user seeking the experience of a premium second home without carrying the full acquisition and operational burden alone.
• The institutional knowledge base is now mechanically deep. Every one of the 795 indexed source paths exists and was read; 682 unique bodies were parsed; 113 repeat reads were avoided through exact-hash mirror inheritance; 4.04 million visible words, 77,356 content units, 53,712 source-grounded claims, 4,354 tables, and 1,052 scripts were captured. This is no longer an index-only assessment.
• Authority, not discovery, remains the limiting factor. Most records still lack owner-ratified currentness. The run routes 564 records to finance, 99 to broader authority owners, 14 to legal, five to prompt-injection quarantine, and 113 to exact-mirror inheritance. No extracted business claim or formula becomes approved merely because it was parsed.
• CoChalet's strongest strategic idea is the integration of four systems that competitors often separate: real-property title, owner-use governance, hospitality operations, and capital formation. Its largest risk is the same integration: an unresolved legal, financial, or operating definition can propagate into owner materials, investor materials, the portal, automation, and brand copy.
• The next competitive advantage is one unmistakable authority chain per role. The correct next move is not broader document creation. It is to name controlling sources, ratify supersession chains, bind quantitative claims to canonical pins, validate the small business-relevant subset of parsed relationships, and make downstream surfaces deny by default when authority is absent.

July 13 Legal-Authority Update

Decision-useful answer: the TCJ record narrows and structures the legal work, but it does not close the launch gate. Counsel states that the revised opinion's conclusions remain unchanged in their existing qualified form. CoChalet therefore has a stronger evidence base for controlled execution, not a new categorical clearance.

What changes in the master analysis

• Legal discovery is no longer the immediate bottleneck. The next step is controlled disposition and implementation of a known legal queue.
• The indivision agreement becomes the principal closure vehicle. Drafting-dependent transfer, use, governance, cost, default, creditor, and exit mechanics should be resolved there rather than through unbounded opinion research.
• Securities characterization remains the highest-consequence legal gate. Owner-user facts, marketing posture, governance reality, promoter dependence, final offering materials, and the selected compliance path must remain aligned.
• Clearance is scope-bound and still open. The 33 yellow passages, final-document review, seven excluded legal domains, securities-path decision, regulator-guidance decision, and dated final opinion remain release gates.

Immediate execution order

1. Complete the 32 substantive yellow-item dispositions; preserve the thirty-third yellow marker for source fidelity.
2. Convert the 15 drafting deferrals into an indivision-agreement issue matrix.
3. Resolve all 50 cross-references to exact controlling sections.
4. Carry the 11 qualified responses into a prohibited-claims and release-constraints register.
5. Obtain counsel review of the final agreement, offering, marketing, sales, and portal surfaces.
6. Close each excluded legal domain and request the final, dated, scope-defined opinion.

Open the complete 169-item legal review workbench.

Source controls: PDF SHA-256 aca45ba8920504798ad53b00c650102efb0d423a085156c246e6a22e48946d1f; audited Markdown SHA-256 b5cbebc58edcb4f07eceab18a3d8d442fe827d4efeda8e3a2ddedf8eea627171; machine audit SHA-256 ab86ee90adf73d2c1c9d65a1d34df1d5c539438b6075313f0caee1de771cbb8b. Status: LOCAL REVIEW / NOT LEGAL ADVICE / NOT COUNSEL CLEARANCE.

Bottom-Line Expert Verdict

CoChalet has progressed beyond a promising concept with scattered files. It has a broad operating thesis, a sophisticated legal-economic design vocabulary, a defined hospitality ambition, extensive brand and product thinking, and unusually complete institutional memory. The work product shows that the team understands the hard parts: title structure, owner-use rights, investor-versus-owner separation, property operations, financing sequence, service delivery, stakeholder communications, and evidence control.

The company is not yet an authority-resolved operating system. The same concepts appear in multiple generations, models, drafts, and stakeholder surfaces. Exact hashes now separate mirrors from true branches, but a large share of those branches still lacks a named controller, effective date, and physical supersession path. CoChalet therefore has more knowledge than it can safely operationalize.

My expert conclusion is precise:

CoChalet's present advantage is the depth of its integrated thesis. Its next advantage must be the discipline with which that thesis becomes one governed operating truth.

What CoChalet Is

The customer proposition

Across product, brand, legal, finance, and operating artifacts, the recurring proposition is consistent at a high level: convert recurring premium vacation-home consumption into a deeded owner-user relationship, then remove most of the coordination burden through professional management and hospitality services.

The product is not simply access to nights. The intended bundle is:

1. A registered or registrable undivided real-property interest.
2. A governed right to personal use under an ownership agreement or convention.
3. A property-level operating system for booking, maintenance, vendors, arrivals, and service requests.
4. A customer interface that makes a legally and operationally complex arrangement feel simple.
5. A controlled resale, transfer, financing, or exit process, subject to the governing legal and financial architecture.

The corpus repeatedly frames the co-owner as an owner-user, not as a passive investor. That distinction is not a marketing flourish. It is a load-bearing design constraint for the product, legal framing, fee structure, portal, owner communications, and sales language.

The property and ownership layer

The source bodies repeatedly describe property-level acquisition vehicles, deeded undivided interests, a convention governing co-use, and sponsor or corporate ownership of any residual interest. The exact entity names, residual percentages, financing relationships, and sequencing vary across source branches and must not be treated as settled from corpus frequency alone.

The stable architectural idea is separation of concerns:

• The property/title layer owns or administers real property.
• Co-owners receive deeded ownership and personal-use rights.
• The management or hospitality layer provides services for fees.
• Investor capital, where used, sits on a separate securities-side architecture from the owner-user proposition.
• Financing vehicles, if implemented, require their own regulatory, credit, security-ranking, tax, and consumer-protection analysis.

This separation is one of CoChalet's most important pieces of institutional knowledge. It protects accounting clarity, legal characterization, stakeholder communication, and operational accountability.

The operating and hospitality layer

The corpus envisions a managed premium experience rather than a bare co-ownership registry. Repeated operating themes include:

• booking and stay allocation;
• arrival and departure coordination;
• property maintenance and issue reporting;
• housekeeping and vendor management;
• concierge and owner services;
• service-level expectations;
• quality assurance and review defense;
• property-level reporting;
• owner communication and governance workflows;
• centralized portal or app access.

The strategic implication is that the product's defensibility will come less from the legal ability to divide ownership and more from consistently delivering a trusted owner experience across properties. The title structure gets the customer into the system; operating excellence determines whether the system earns referrals, renewals, resale liquidity, and brand trust.

The capital and financing layer

The corpus contains several capital architectures for bridging property acquisition, admitting co-owners, retaining a residual property position, financing operations, and potentially financing individual co-owner purchases. These materials show sophisticated awareness of capital seniority, return-of-capital mechanics, private-placement boundaries, debt security, owner financing, and the need to keep investor economics separate from co-owner language.

The authoritative details remain gated. This report therefore does not restate transaction values, rates, returns, valuations, ownership percentages, or timelines as current truth. Those belong only in CFO-controlled pins and owner-ratified transaction documents.

The durable insight is that CoChalet is not a simple software-capital model. It is a capital-intensive operating business whose early execution depends on matching the acquisition sequence, co-owner conversion sequence, mortgage or hypothec structure, investor expectations, and operating ramp without allowing one stakeholder's economics to contaminate another's disclosures.

The legal and regulatory layer

The source bodies repeatedly identify Quebec civil-law co-ownership as a core structural differentiator. They also reveal the breadth of the legal perimeter:

• indivision and registered title;
• conventions governing use and co-owner obligations;
• securities treatment of investor-side instruments;
• the intended non-security treatment of deeded owner-user interests;
• real-estate brokerage triggers;
• mortgage or hypothec registration and priority;
• consumer-credit and disclosure rules if owner financing is offered;
• tax characterization at property, entity, owner, lender, and investor levels;
• privacy and data handling;
• French-language and consumer-facing communications;
• trademark, copyright, and trade-secret protection.

The corpus is strongest when it treats legal review as a product-design input rather than a final disclaimer exercise. It is weakest when a draft or marketing surface states a legal conclusion before the controlling opinion, factual predicates, and audience scope are explicit.

The brand and market position

The brand system consistently seeks a premium alpine identity built around ownership, belonging, ease, and managed hospitality. Repeated motifs include a dusk-to-dawn visual journey, deep alpine tones, warm interior light, and a simple progression from discovery to ownership. The corpus also contains several candidate slogans, product tier names, and economic comparisons.

The brand challenge is not lack of ideas. It is claim discipline. Some high-impact copy compresses the concept effectively but risks implying appreciation, investment return, guaranteed economic advantage, or operational maturity beyond what the current gates authorize. The correct brand system must preserve aspiration while remaining exact about current operations, legal character, and quantitative evidence.

The portal and technology layer

The technology vision is a role-aware operating interface connecting owners, property operations, finance, documents, requests, and service data. The corpus contains portal concepts, app flows, automation specifications, data-feed designs, dashboards, command centers, and presentation tooling.

The core product insight is sound: the customer should not have to understand the organizational complexity behind a stay, a fee, a service request, a document, or a property update. The portal should compile that complexity into clear owner actions and evidence-backed status.

The governance risk is equally clear. A current-looking portal can become the most powerful misinformation surface in the company if its numbers, status labels, documents, or automation rules are not bound to current controlling sources. Portal convenience must be downstream of authority, not a substitute for it.

What Is Known, What Is Observed, and What Is Not Yet Authorized

Evidence Base: What V2 Actually Adds

The V1 report analyzed the HTML compendium as an index. V2 incorporates the completed source-body authority-resolution run.

All deterministic and mechanical checks in the independent audit pass. This proves population closure, source readback, parser evidence, exact-hash inheritance, packet seals, formula expression hashes, required formula fields, control hashes, and report rendering. It does not prove business authority.

Interpretation: source discovery is closed to the indexed denominator. The remaining work is not finding the 795 files; it is deciding which bodies control and which must remain historical, quarantined, or derivative.

V1-to-V2 Maturity Shift

Scale: 0 absent, 1 blocked scaffold, 2 partial, 3 usable but gated, 4 controlled, 5 mechanically verified. The scores are an ordinal planning heuristic, not a probability. Open authority gates override the total.

V1 heuristic total: 27/75.
V2 heuristic total: 54/75.

The doubled score is meaningful but easy to misread. Mechanical knowledge maturity improved sharply. Decision authority did not improve at the same rate. CoChalet now knows much more precisely what it has, where it came from, what it says, and where it conflicts. It still needs named humans to decide what binds.

Level 1: Evidence Integrity

Maturity: 5/5, mechanically verified.

The source index is hash-locked, every indexed path was rehashed, every packet and body digest was sealed, and formula expression hashes were independently rechecked. The authority-resolution artifact set itself was hashed into a stable manifest across two observations.

The expert lesson is that CoChalet's strongest reusable capability is evidence discipline. This should become standard operating infrastructure: every current-canon release, workbook population, legal package, portal build, or investor deck should carry source hashes, controller hashes, parser versions, release gates, and readback proof.

Remaining limit: evidence integrity proves what was reviewed. It does not authorize the reviewed statement.

Level 2: Population Completeness

Maturity: 5/5, closed to the 795-record denominator.

All 795 indexed paths exist, are unique, and were read. No indexed record is missing, inaccessible, or unsupported. Exact mirrors were handled through content hashes, so no source silently disappeared when duplicate reads were avoided.

This closes a major epistemic gap from V1. The compendium is no longer merely a routing map for hypothetical source reading; it is the control population for a completed source-body pass.

Remaining limit: the 795 records are the exported denominator, not proof that every relevant CoChalet file everywhere is represented. Future refreshes need delta detection and explicit out-of-index intake.

Level 3: Representation Fidelity

Maturity: 4/5, controlled parsing.

V2 decomposes source bodies into paragraphs, headings, lists, tables, code, scripts, captions, and other content units. Embedded JSON was parsed where present. Directly referenced local Markdown, JavaScript, JSON, PDF, and workbook bodies were parsed separately. PDF pages were rendered for visual inspection, and the workbook was inspected at sheet and cell level.

This means the system can now distinguish authored prose from tables, interface code, native dependencies, and repeated support content. That is a decisive improvement over file-size or record-count heuristics.

Remaining limit: 511 direct references were missing or nonfile references. Many may be irrelevant support links, but materiality has not been adjudicated. One authored inline script is syntactically malformed and remains an explicit exception.

Level 4: Corpus Topology

Maturity: 5/5, exact-hash topology resolved.

The 795 paths map to 682 unique bodies. Eighty-two exact-mirror groups contain 195 records, allowing 113 reads to inherit sealed evidence without semantic duplication. Same-name, different-body sources remain separate.

This resolves the most important identity error in the old system: path uniqueness is not knowledge uniqueness. CoChalet can now preserve every storage location while reasoning over unique content.

Strategic implication: future refreshes should be content-hash first. New paths with existing hashes are mirror events; new hashes are review events.

Level 5: Domain Coverage

Maturity: 4/5, body-grounded but not owner-ratified.

The source bodies contain substantial material across finance, legal and governance, operations, product and portal, marketing and brand, technical systems, and mixed operating records. Claims were extracted from actual content units rather than inferred solely from filenames.

The corpus is notably cross-functional. Finance concepts appear in brand and product surfaces; legal concepts appear in owner and investor materials; operating promises appear in sales and portal artifacts. That interdependence makes role-aware impact analysis essential.

Remaining limit: topic organization is useful for routing, not authority. A finance-related statement in a marketing file does not become finance-approved because a classifier recognized it.

Level 6: Provenance and Chain of Custody

Maturity: 3/5, usable but incomplete.

V2 adds content hashes, source paths, exact mirror groups, parser evidence, route context, and source-level review packets. This is materially stronger than model-inferred authorship labels.

Observed provenance and inferred provenance still need strict separation. A model may suggest that writing resembles a specific agent or generation, but only filesystem, Git, generator, route, author, and release evidence can establish chain of custody.

Required next control: every controlling source should record owner, ratifier, effective date, generator or author, release event, and supersession decision in addition to its content hash.

Level 7: Identity, Duplication, and Mirrors

Maturity: 5/5 for exact identity; semantic version identity remains owner-gated.

Exact content identity is resolved. The system can identify byte-identical mirrors and avoid repeated semantic reads while preserving every path. Divergent bodies remain distinct even when names match.

The remaining identity task is semantic: deciding whether two divergent bodies are successors, scoped variants, drafts, stakeholder editions, or unrelated artifacts. That decision cannot be made from similarity or recency alone.

Required identity rule: content hash + semantic family + authority tier + effective period + supersession chain.

Level 8: Temporal and Version Semantics

Maturity: 3/5, signals routed.

V2 assigns bounded lifecycle signals from source evidence and paths: archived, draft, hold, superseded, or unresolved currentness. It does not silently treat dates or directory names as ratification.

Observed population:

• unresolved currentness: 723;
• archive signal: 14;
• draft signal: 25;
• hold signal: 8;
• superseded signal: 25.

Interpretation: the system can find probable lifecycle edges, but most records still lack a controlling effective-state decision.

Level 9: Lifecycle State

Maturity: 2/5, owner gate open.

V1 had only extracted. V2 distinguishes observed lifecycle signals and routes every record to a terminal review disposition. That is progress, but a signal is not a ratified state.

The target lifecycle remains:

DRAFT -> REVIEW -> RATIFIED -> CURRENT -> SUPERSEDED -> ARCHIVED

Parallel safety states should include HOLD, QUARANTINED, and PROHIBITED. Every transition needs an actor, timestamp, reason, evidence, affected audiences, and successor pointer.

Level 10: Claims and Authority Semantics

Maturity: 3/5, claims extracted and authority unresolved.

The run created 53,712 source-grounded claims with source units and statement hashes. Claims are classified across finance, legal, ownership or investor, product, operations, automation, marketing, lifecycle, and authority themes.

This proves what source bodies state. It does not prove that the statements are true, current, or approved. The staged role-aware front door intentionally reports zero approved claims and zero approved formulas until named owners ratify controllers.

This distinction is the heart of information mastery:

• Observed claim: a source states it.
• Derived claim: a controlled transformation produces it.
• Authority claim: a named owner ratifies it for a defined scope.
• Published claim: an authority claim is released to an approved audience.

CoChalet should never skip from observed to published.

Level 11: Contradictions and Reconciliation

Maturity: 3/5, routed but not adjudicated.

The contradiction and reconciliation register contains 535 rows. It combines same-name divergent-body branches with canon gates for unpinned, retired, or prohibited quantitative evidence.

The system correctly refuses to choose a winner by filename, polish, recency, or model confidence. The remaining work belongs to CFO, legal, founder, operations, product, engineering, and CMO according to impact.

Expert warning: contradiction volume is not merely archival debt. Each unresolved branch can feed multiple downstream surfaces. Resolution should therefore be impact-weighted, not alphabetical.

Level 12: Formula and Dependency Mastery

Maturity: 3/5, syntax proven and business meaning gated.

V2 contains 107,869 formula or relationship rows:

• 38,159 parser-backed inline JavaScript relationships;
• 69,040 parser-backed native JavaScript relationships;
• 670 source-grounded prose relationship candidates.

Most executable relationships are interface or vendor code. The system routes 104,914 as nonauthoritative code relationships and 2,955 as business-domain signals requiring semantic and authority review.

The business-signal queue spans capital, cost, operations, ownership, return, risk, and valuation concepts. Parser evidence proves syntax, identifiers, conditions, source spans, and dependency edges. It does not prove correct units, business definitions, current assumptions, legal permissibility, or authorized outputs.

Required next control: map every material business relationship to owner-controlled inputs, outputs, units, exceptions, tests, canonical pins, and approved stakeholder surfaces.

Level 13: Risk and Model-Control Reliability

Maturity: 4/5, deterministic controls pass.

The source-body program corrected the main weakness identified in V1. Local models were not trusted to clear sources, confirm provenance, validate formulas, or choose canon. Mechanical controls handled hashes, paths, parser structure, mirror inheritance, source spans, retired-claim matching, prompt-injection quarantine, and completeness.

Five bodies containing prompt-injection indicators were quarantined. Potential secret indicators were treated as safety signals and were not reproduced into this report. A 16-test adversarial harness passed after three defects in the test fixtures themselves were corrected.

Remaining limit: deterministic checks can prove consistency and detect classes of risk. They cannot ratify business meaning or legal sufficiency.

Level 14: Retrieval and Governance Architecture

Maturity: 3/5, role-aware front door staged.

The staged authority front door separates founder, finance, legal, operations, product, engineering, marketing, owner, investor, and public-surface concerns. It identifies available controllers, candidate populations, prohibited downstream uses, and next required decisions.

This is the right architecture. The front door remains staged because most roles have not selected a unique controlling source set and supersession chain.

The final retrieval contract should return, in order:

1. controlling source;
2. owner and ratifier;
3. effective date and lifecycle state;
4. exact claim or formula span;
5. canonical inputs and derivation lineage;
6. approved stakeholder surfaces;
7. superseded alternatives;
8. unresolved holds and exceptions.

Level 15: Decision Readiness

Maturity: 2/5, staged only.

CoChalet can now support evidence-grade corpus navigation, exact-mirror handling, source-body search, source-grounded claim review, formula-syntax review, native-dependency triage, stakeholder impact mapping, and owner-specific decision queues.

It still cannot support automatic current-canon selection, public claims, owner or investor disclosures, legal conclusions, formula implementation, or production promotion from the corpus alone.

The decision-readiness blocker is not lack of analysis. It is the absence of completed owner ratification at the final mile.

Interpretation: the corpus is completely dispositioned but not broadly approved. Population closure and release authority are different metrics.

Cross-System Expert Findings

1. CoChalet's real moat is orchestration

The legal structure, property operations, hospitality experience, capital sequencing, brand promise, and portal must agree. A competitor can copy a landing page or a co-ownership calculator. It is harder to copy a governed system that makes deeded ownership, co-use, service delivery, financing, and customer trust work together.

2. The first operational proof matters more than a larger strategy library

The corpus contains extensive future-state thinking. The highest-value evidence now is operational: one controlled property acquisition, one legally clean co-owner close, one repeatable booking and service cycle, one reconciled owner statement, one issue-resolution history, and one portal path whose data traces to controlling sources.

3. The owner-user and investor stories must remain physically separate

The corpus understands this conceptually, but document branches create leakage risk. The owner-user story is about use, title, obligations, service, and the ordinary incidents of ownership. The investor story is about securities-side capital, risk, governance, and permitted returns. Shared numbers or vocabulary should flow only through controlled, audience-aware compilation.

4. Hospitality quality is the economic engine of trust

CoChalet's brand promise depends on reliable execution: clean properties, predictable arrivals, responsive service, transparent costs, fair allocation, and visible issue resolution. The best legal structure cannot compensate for unreliable owner experience.

5. The portal should be a compiler, not a content warehouse

The portal should display governed outputs from current sources. It should not become another place where numbers and definitions are manually copied. Every displayed figure, status, policy, and document should carry a source anchor, freshness state, and release rule.

6. The knowledge system itself can become an operating advantage

Few early-stage real-estate platforms have this volume of structured institutional memory. If CoChalet converts it into one owner-governed authority graph, it can reduce diligence friction, onboarding time, compliance drift, and cross-functional rework.

Principal Risks

P0: Authority drift

Multiple polished versions can look current. A wrong winner can contaminate legal, finance, owner, investor, product, or public surfaces.

Control: one named controller and physical supersession per semantic family.

P0: Quantitative leakage

The source bodies contain many financial literals and relationship candidates. Exact matches to canonical values still may be wrong for the context, and unpinned values may be stale or hypothetical.

Control: CFO pin binding, retired-claim deny rules, audience-specific compilation, and no naked financial literals in governed downstream sources.

P0: Securities and owner-language contamination

Investor concepts can leak into co-owner messaging, or owner-use concepts can be used to oversimplify securities-side risk.

Control: separate templates, separate approval chains, and blocked vocabulary by stakeholder surface.

P1: Operational overstatement

Future-state hospitality, app, service, and scale concepts may be described as current capability.

Control: explicit operations mode such as DESIGNED, PILOT, or LIVE, with evidence-backed transitions.

P1: Portal authority failure

A well-designed portal can create false confidence if its data and documents are not current.

Control: source anchors, freshness checks, deny-by-default rendering, and visible holds.

P1: Unresolved native dependencies

Some referenced native artifacts are unavailable or nonfile references. Most may be immaterial, but that has not been proven.

Control: classify all 511 references by business materiality before claiming complete native closure.

P2: Local-model false confidence

Organization models can produce neat classifications that exceed their evidence.

Control: confine them to bounded organization tasks, require abstention, and score critical false-clears separately from average accuracy.

The Correct Execution Program

Phase 1: Ratify the authority graph

1. Founder names the final owner and ratifier for each domain.
2. Each owner selects one controlling source set and one current front door.
3. Divergent branches receive SUPERSEDE, COEXIST_BY_SCOPE, HOLD, or ARCHIVE decisions.
4. Current sources receive effective dates, audience scope, and physical successor links.

Phase 2: Close high-consequence finance and legal gates

1. CFO adjudicates unpinned and retired quantitative evidence by decision impact.
2. Legal identifies the binding owner, investor, property, brokerage, financing, privacy, and language authorities.
3. Founder, CFO, and legal jointly approve the boundary between owner-user and investor-facing language.
4. No portal, deck, owner document, or public copy compiles from unresolved sources.

Phase 3: Convert formulas into governed business logic

1. Reduce the 2,955 business-signal relationships to material operating formulas.
2. Assign owner, inputs, outputs, units, conditions, exceptions, and tests.
3. Bind finance inputs to canonical pins.
4. Prohibit derived values from being manually restated downstream.
5. Publish regression checks for every owner, investor, operations, and portal surface.

Phase 4: Prove the operating loop

1. Select the smallest real property and owner journey that exercises the full system.
2. Verify acquisition, closing, booking, service, issue, reporting, and governance events.
3. Capture evidence and exceptions without turning pilot behavior into unqualified scale claims.
4. Promote operating states only when the evidence exists.

Phase 5: Make retrieval role-aware and deny by default

1. Founder sees strategic decisions and open owner gates.
2. Finance sees pins, formulas, reconciliations, and retired claims.
3. Legal sees binding authorities, factual predicates, and affected surfaces.
4. Operations sees live procedures and service evidence.
5. Product and engineering see current data contracts and executable specifications.
6. Marketing sees approved claims and prohibited language.
7. Owners and investors see only their authorized, audience-specific outputs.

Ninety-Day Decision Sequence

Days 1 to 15: Controller selection

• Name domain owners and ratifiers.
• Resolve the highest-impact branch families.
• Freeze current pointers for finance, legal, operations, portal, and external claims.
• Publish the first approved role-aware front door.

Days 16 to 35: Quantitative and legal reconciliation

• Close the highest-consequence CFO queue.
• Convert legal open questions into written rulings or explicit holds.
• Recompile owner, investor, and portal specifications from approved inputs.
• Add automated retired-claim and naked-number gates.

Days 36 to 60: Formula and interface control

• Ratify material business formulas.
• Bind them to canonical pins and source-controlled tests.
• Define portal freshness and authority contracts.
• Prove that downstream values are compiled, not copied.

Days 61 to 90: Operating proof

• Execute one end-to-end owner and property workflow.
• Measure service delivery, exceptions, owner clarity, and data lineage.
• Update the authority graph from observed operating evidence.
• Decide which claims may graduate from designed to pilot or live.

Owner Decisions Required

1. Which entity and agreement architecture controls the initial property and co-owner closing?
2. Which owner-facing rights, obligations, services, and transfer rules are binding?
3. Which investor-side instrument and disclosures are approved, and which internal scenarios must never appear externally?
4. Which revenue, fee, cost, reserve, and service definitions are current CFO canon?
5. Which legal opinions or written rulings are still required before customer or investor use?
6. What is the current operating state: designed, pilot, or live, by capability?
7. Which portal specification and data contract are controlling?
8. Which brand claims are approved for public, owner, and investor audiences?
9. Which semantic source families may coexist by audience rather than supersede one another?
10. What evidence is required to promote from staged to production?

Further Questions

• What is the minimum owner experience that must be flawless before adding another property?
• Which operating tasks create the greatest risk of service failure or owner distrust?
• Which legal-economic choices are true competitive advantages, and which add complexity without sufficient customer value?
• Which data should an owner see directly, and which requires explanatory context or professional review?
• Which quantitative claims actually improve conversion, and which create avoidable legal or trust risk?
• Which parts of the premium experience are operationally distinctive versus merely aesthetic?
• What is the smallest set of current sources from which every approved surface can be compiled?

Verification and Evidence

Primary V2 evidence:

• 13_SOURCE_BODY_COVERAGE_AND_EXCEPTION_AUDIT.json
• 14_COCHALET_SOURCE_BODY_MASTER_SYNTHESIS.md
• 12_ROLE_AWARE_AUTHORITY_FRONT_DOOR.json
• 01_INDEXED_SOURCE_MANIFEST.jsonl
• 02_CONTENT_HASH_MIRROR_REGISTRY.jsonl
• 05_CONTENT_UNIT_LEDGER.jsonl
• 06_SOURCE_GROUNDED_CLAIM_LEDGER.jsonl
• 07_PARSER_BACKED_FORMULA_LEDGER.jsonl
• 08_AUTHORITY_LIFECYCLE_REGISTER.jsonl
• 09_CONTRADICTION_AND_RECONCILIATION_REGISTER.jsonl
• 10_NATIVE_DEPENDENCY_POPULATION.jsonl
• 11_STAKEHOLDER_SURFACE_IMPACT_REGISTER.jsonl
• 15_ARTIFACT_MANIFEST_SHA256.json

Mechanical verification includes fresh source rehashing, packet and body seals, claim and formula hash checks, mirror inheritance checks, native-source seals, control hashes, a 16-test adversarial harness, HTML parsing, and desktop/mobile browser review.

Caveats and Assumptions

• This report separates corpus observation from authority. Repeated statements are not automatically current truth.
• No real transaction values, credentials, secrets, or sensitive personal data are reproduced here.
• The maturity score is a planning rubric, not a probability or company valuation.
• The 795-record denominator is complete for the locked compendium export, not a claim that every relevant file outside that export was discovered.
• Parser-backed syntax does not establish business validity, legal sufficiency, or approved formula meaning.
• Exact canonical literal matches still require context and stakeholder-surface authority.
• Local model outputs were confined to organization-only evaluation and were not used as authority evidence.
• The requested Sol 5.6 Ultra model identity was not independently verifiable and is not claimed.
• No source corpus, production system, public surface, or current-canon file was mutated.
• Final state remains STAGED / VERIFIED-TO-MECHANICAL-SCOPE / NOT PROMOTED.

Final Expert Recommendation

CoChalet should treat the corpus as a nearly complete map of institutional memory and an incomplete map of institutional authority. The company does not need more ideas before it can move. It needs fewer competing truths.

The winning operating pattern is:

1. one controller per role;
2. one physical supersession chain per semantic family;
3. one canonical pin source for each governed number;
4. one owner-ratified formula definition for each material calculation;
5. one compiled, audience-specific output path;
6. one evidence-bearing promotion gate.

When those controls are in place, CoChalet's unusually rich institutional knowledge can become an execution advantage rather than a source of ambiguity. Until then, the correct verdict is REVISE, and the correct posture is staged, controlled, and honest.