- Version:
V42 - V42 state: canonical promotion complete; V42 notes record accepted Depositing, Reading, Need review, Finding Fits, settlement, delivery, AI-reading demonstration, local/staging rehearsal, and promotion-readiness evidence
- Current canonical/latest target:
V42 - Canonical proof-source commit:
5c9c0270b9d864fe13b7e0a429700e1c9a7689d9 - Prior canonical anchor:
BITCODE_SPEC_V41.md - Prior generated proof appendix:
BITCODE_SPEC_V41_PROVEN.md - Generated structured artifact inventory: active canonical
.proofs/v42/spec-family-report.json,.proofs/v42/canonical-input-report.json,.proofs/v42/canon-posture-drift-report.json,.proofs/v42/depositing-shortest-path.json,.proofs/v42/reading-shortest-path-state-machine.json,.proofs/v42/readneed-review-resynthesis-product-closure.json,.proofs/v42/readfitsfinding-preview-quote.json,.proofs/v42/settlement-rights-delivery.json,.proofs/v42/ai-reading-demonstration.json,.proofs/v42/local-staging-mvp-rehearsal.json,.proofs/v42/promotion-readiness-report.json, V42 gate-quality and promotion workflow evidence, andBITCODE_SPEC_V42_PROVEN.mdas the generated proof appendix for V42 promotion - Source parity state: V42 source-side Depositing shortest path, Reading shortest path state machine, ReadNeed review/resynthesis closure, ReadFitsFinding preview and quote closure, settlement rights delivery, AI-reading demonstration, local/staging-testnet rehearsal, workflow, and promotion surfaces are canonicalized in the promoted V42 file family
- Scope: V42 canonical notes for reliable MVP experience over promoted V41 prompt-program excellence canon
- Last fully realized canonical target preserved in source:
V42
This notes companion records the working V42 product plan and simplified reading.
It does not override BITCODE_SPEC_V42.md.
Product behavior changes remain blocked until the relevant V42 gate admits and verifies them.
Gate 1 is intentionally specification, documentation, workflow, and checker posture only. It opens the reliable MVP experience version after V41 prompt-program promotion and prepares later gates to implement Depositing, Reading, settlement, delivery, and demonstration behavior without ad hoc scope drift.
The V42 Depositing path should minimize the journey from source material to Depository admission proof.
The user needs to know that the source is admitted, searchable for future Need-Fit work, and eligible for BTC compensation if it contributes to a synthesized AssetPack.
The UX can stay simple, but expandable details must show source authority, admission proof, storage projection, search-document posture, compensation route, and repair state.
Gate 2 implements that note through DepositorySupplyCompensationPreview, deposit route depositoryEvidence.compensationPreview, Terminal compensation readback rows, and .proofs/v42/depositing-shortest-path.json.
Local/staging rehearsal for this gate means the deposit route, source-safe evidence projection, generated artifact, package test, and protocol route test can run without value-bearing mainnet behavior.
The staging-testnet lane may read the same source-safe roots and ledger keys, but secrets and private source remain outside generated artifacts.
The V42 Reading path should be a five-step enterprise flow:
- request read;
- review synthesized Need;
- request Finding Fits;
- review source-safe AssetPack measurements and preview metadata;
- buy, settle, transfer rights, and receive repository delivery.
The default UI should be guided and low-detail.
The rich execution log, proof roots, telemetry rows, and ledger/storage details remain available on expansion.
Gate 3 implements the current V42 state-machine layer with TerminalEnterpriseReadingRouteState.
The recoverable route state includes transaction id presence, readingStage, active-stage hydration, retry and restart posture, source-safe failure kind, and repair actions.
This makes refresh, route handoff, and recovery inspectable without disclosing protected source, protected prompts, raw provider responses, unpaid AssetPack source, wallet private material, private settlement payloads, or ledger authority.
Gate 4 closes the ReadNeed product loop.
ReadNeedComprehensionSynthesis now has a V42 proof target for source-safe Read Request persistence, synthesized Need storage, feedback and resynthesis lineage, Need measurement roots, accepted Need admission, rejected Need posture, runtime storage projection, and telemetry receipts.
The Terminal readback must show the Need, review state, runtime root, storage root, telemetry root, blockers, PTRR step identity, and storage record roots without exposing protected source, raw protected prompts, raw provider responses, unpaid AssetPack source, wallet private material, or private settlement payloads.
Finding Fits remains blocked until the Need is accepted.
Rejecting a Need must preserve feedback and keep the user on the review/resynthesis step.
V42 must make the preview valuable without leaking the source-bearing AssetPack before settlement. Readers may see measurements, fit confidence, quote posture, selected-fit provenance summaries, proof roots, and source-safe explanations. They may not see protected source, raw provider responses, protected prompts, private settlement payloads, wallet private material, or unpaid AssetPack source before BTC settlement and BTD rights transfer.
Gate 5 closes the source-safe Finding Fits preview boundary.
ReadFitsFindingSynthesis must already have an accepted Need, then search many Depository candidates across lexical, symbolic, path, metadata, measurement, embedding-vector, and provider-specific channels.
Selected-fit provenance carries selected candidate ids, fit deposit ids, query and ranking roots, proof roots, measurement roots, and reconciliation readback roots.
AssetPackPreviewBoundary is the review object before payment: it contains the source-safe preview, deterministic quote receipt, disclosure review, settlement instructions, delivery lock, storage projection, replay receipt, and repair posture.
The route and Terminal readback expose those fields as metadata only; protected source and unpaid AssetPack source remain absent until Gate 6 settlement, rights transfer, and delivery unlock.
Gate 6 closes the paid settlement and source-bearing delivery boundary.
AssetPackSettlementRightsDeliveryBoundary is the reviewable post-payment object: it records BTC payment observation, confirmed finality, source-to-shares compensation, BTD read and rights transfer receipts, settlement unlock, source-bearing pull-request delivery unlock, ledger/database/object-storage reconciliation, storage projection, replay receipt, and repair posture.
The live Vercel Sandbox harness materializes this boundary after ledger readback, the route summarizes it as source-safe metadata, and Terminal renders settlement rights, source-to-shares compensation, delivery unlock, reconciliation, and replay roots.
Protected source, private settlement payloads, wallet private material, raw protected prompts, raw provider responses, and credentials remain absent from the generated artifact, route summaries, and pre-delivery readback.
The standalone demonstration should prove why Bitcode matters for AI-dominant Reading.
A deposited proprietary or otherwise non-public technical intelligence source should contribute to an AssetPack that measurably improves an AI system's training, prompt/context, or evaluation result beyond what a public-data-only baseline can do.
The demonstration must remain minimal, local, deterministic where feasible, and self-contained inside protocol-demonstration/.
Gate 7 records that proof as a public-data-only baseline, a reviewed local Need, a local Depository Finding Fits step, a selected AssetPack preview, an AssetPack-enhanced AI-reading answer, and deterministic benchmark uplift.
The demonstration proves value without weakening the settlement-gated visibility boundary: protected source is still withheld until settlement, and the demonstration remains independent from product runtime imports.
Gate 8 records the source-safe full MVP rehearsal posture across local and staging-testnet lanes.
It composes the already-closed V42 Depositing, Reading state, ReadNeed review/resynthesis, Finding Fits preview/quote, settlement rights delivery, and AI-reading demonstration proofs into one operator-readable rehearsal artifact.
The rehearsal binds ReadingLocalStagingRehearsal, Vercel Sandbox operator receipts, staging-testnet Supabase project tkpyosihuouusyaxtbau, rich execution-log readback, database streaming, ledger/database/storage reconciliation, post-settlement pull-request delivery, and explicit value-bearing mainnet blocking.
It is intentionally source-safe metadata only: secrets, protected source, raw protected prompts, raw interpolated prompts, raw provider responses, unpaid AssetPack source, wallet private material, private settlement payloads, and live rehearsal logs remain outside tracked artifacts.
Gate 8 does not implement the V43+ route vocabulary, but it keeps the future product shape visible: /read, /deposit, and /packs remain the next route cleanup once V42 proves the current MVP path.
Gate 9 records the V42 promotion-readiness closure state.
The generated artifact .proofs/v42/promotion-readiness-report.json binds every V42 product proof from Gates 2 through 8, the V42 promotion workflow, generated proof appendix support, promotion dry-run support, source-safety checks, and value-bearing mainnet blocking.
The intended post-promotion runtime posture is active V42 / draft V43, so V43 can start from the reliable MVP experience canon and focus on the next product vocabulary and deposit-side evolution.
V43 or a later explicitly opened version should evolve the deposit side into an agentic AssetPack option experience for enterprises.
Repository-installed Bitcode Agents should compare a connected enterprise codebase, the current Bitcode Depository, and Reading activity to propose deposit AssetPack options.
Those options should be source-safe, sub-critical, likely positive ROI, and approve/rejectable before Depository admission.
That later version should split /terminal into /read and /deposit, and rename /exchange to /packs across routes, code naming, docs, and operator vocabulary.
The route model should be AssetPacks in and AssetPacks out: /deposit creates reviewable deposit AssetPack options from connected source, depositor instructions, and Bitcode's observed Needs; /read creates reviewed Need-Fit AssetPack previews and only unlocks source-bearing delivery after settlement; /packs is the searchable master-detail activity route for deposited packs, settled read packs, compensation posture, quotes, rights transfer, delivery, and repair.
The /packs master view should support column sorting, filtering, and search over measurements, synthesized AssetPack titles and descriptions, values, activity or transaction type, settlement posture, and compensation state.
The detail view should expose the selected activity's source-safe data, proof roots, telemetry, ledger/database synchronization, and expandable payloads without replacing the short default path.
Outside public documentation, product UX should avoid self-referential explanatory copy; route structure, concise labels, progressive detail, and proof-on-expand must make Depositing, Reading, and Pack activity self-explanatory.
The current transitional Terminal/Exchange UX remains too dense for the final MVP posture.
V43+ route vocabulary must remove Exchange naming from route and component prefixes in favor of Packs, split Terminal into /read and /deposit, keep /packs as searchable master-detail activity, and use rich themed reusable components without relying on in-product explanatory copy to compensate for unclear flows.
Bitcode is active at V41. V42 drafts the next canon: the reliable MVP experience for enterprise Depositing and Reading. The central product promise is that a user can deposit source, request a read, review a synthesized Need, find fitting Depository sources, preview a source-safe AssetPack, pay in BTC/BTD settlement terms, receive rights, and get repository delivery.
Read V42 as shortest credible paths over existing protocol law. If a product step cannot be routed, stored, replayed, telemetered, proven, and source-safely explained, it is not V42-ready.
- Do not implement product behavior in Gate 1.
- Do not split
/terminalor rename/exchange. - Do not expose protected source or unpaid AssetPack source.
- Do not bypass Need review before Finding Fits.
- Do not claim settlement, BTD rights transfer, or repository delivery without synchronized ledger/database/storage proof.