Gå til hovedindholdet

Open Manuscript Studio 1.0 — Executive Summary

Research-project contract reconciliation — 2026-11-15​

The research-project format drift is repaired through one canonical Draft envelope, compatibility migration and independent schema/golden tests. Project identity and opaque data survive graph edits, local reload and undo. This is a bounded local/manual-transfer slice of ADR-036; direct cloud synchronization, conflict receipts and searchable publication are still planned capabilities.

Status: proposed coordinated repair pending companion merges/CI; deployment and provider acceptance remain separate. No manuscript/container generation change or database migration.

Contract, compatibility and current implementation.

Text-box and research provenance — 2026-11-13​

The graphical-text and research-diagram additions need a coordinated contract repair. Canonical Draft schemas and an independent experimental text-box profile retain editability and recovery; reviewed graph snapshots carry visible/structured external provenance or a private author-declared own-work marker. This is not a stable-format promotion, review certification or completed release acceptance.

Status: proposed coordinated repair; merge/CI and deployment acceptance remain separate. No database migration, container-generation change or new review authority.

Profile, decision and acceptance scope.

Privacy/API boundary repair — 2026-11-08​

The proposed privacy/account-copy repair keeps useful operator notices and downloads while restoring additive /api/v1/privacy and /api/v1/account/data. Legacy routes are shared-handler facades with deprecation signals and one shared quota. Strict Zod DTOs generate the checked OpenAPI contract; an injected application service owns scope/serialization and identity repository adapters own Prisma. The copy remains explicitly partial, never a complete rights response or manuscript backup.

The selected server operator owns reviewed controller facts; this does not confer account, editorial or review authority. Existing identity/API/privacy CI and exact-candidate release gates still apply. Deploy API before clients; no OMI change, new migration or authoring restriction is introduced. Source repair is proposed until the companion PR merges and passes CI; deployment and legal acceptance remain separate.

Studio ADR-034; Studio PR #886.

PKP support and release evidence — 2026-11-07​

OJS 3.5.0-4/-5/-6 are in the Stable acceptance scope; OMP 3.5.0-4/-5/-6 remain Preview. All six named E02 jobs/archives are mandatory regression evidence. Add ojs-3.5.0-6 to the current v1 receipt (14 manual gates); preserve historical 13-gate receipts without treating them as approval for the expanded scope.

This enables the new compatibility scope without promoting OMP to Stable or approving an RC/deployed upgrade. Current support decision.

Publication rendition boundary — 2026-11-04​

Publication layout is a downstream editable version, not source processing. Separate frame text is legitimate; the earlier categorical drift finding is narrowed to unbound source lineage and unproved semantic fidelity/assurance. The author may edit every part. Name and ORCID can provide visible credit; internal source revision/digest and rendition identity carry provenance without printing technical identifiers. Existing RC and review gates remain unchanged.

Inspected source, decision and acceptance work.

Outbound security and acceptance checkpoint — 2026-11-02​

Keep the working security implementation. The change needed now is explicit compatibility documentation and traceable release evidence: closed security issues are recorded as closed, while actual host acceptance and exact-candidate receipts remain separately assessed. No schema/API freeze, release promotion, database migration or manuscript editing restriction follows from this reconciliation.

Current evidence authority: reviewed security/acceptance checkpoint. Its dated status supersedes older claims below that #765/#774 are open or that connection-time pinning has no implementation. Historical findings remain historical; closed issues do not supply missing RC receipts.

Historical checkpoint: 2026-11-01. Earlier checkpoints are historical where superseded. Proposed repairs require merge/CI; real acceptance remains separate.

Session, integration and release evidence reconciliation — 2026-11-01​

Reviewable repairs: Studio #813 · Studio #814 · Studio #815 · Studio #816 · Studio #817. During the repair, Studio #812 merged the E02 collector. #817 reuses that existing latest-attempt/same-SHA/job/archive gate (including OMI-SPEC-320 conformance), adds candidate-only build/review and binds actual publication bytes/manual receipt; it does not introduce a competing workflow inventory.

Keep the existing editor, immutable revisions, container parser and renderers. The repair is bounded and proposed until companion merge/CI; it does not establish 1.0 readiness.

The verified highest-risk defects are the API-origin Office sandbox privilege and the global native working-file path. Use an opaque-origin sandbox and per-manuscript save bindings. An application-owned document-session manager preserves pending batches, selection and recovery through multiple opens; checkpoint-on-switch already existed, so the earlier claim of automatic pending loss on that path is corrected.

A shared synthetic container fixture exposed a subtler integrity problem: deleting previews inside historical nested sections can change the existing v1 state digest. Preserve exact historical snapshots, declare the current-document serialization profile and verify the same golden in both repositories. Do not silently rewrite digest semantics or create a large migration framework.

The extension registry receives an additive /api/v1 facade over a shared router/service/repository; it still grants no manuscript access. Production PDF requires the isolated worker; missing host configuration fails PDF only. For 1.0 publication, bind owner acceptance to the actual prior native artifacts and hashes plus same-SHA CI/CodeQL/PKP/readiness results and manual host/device evidence. Missing evidence blocks tagging, not authoring.

No app release bump, DB migration, new OMI schema generation or OJS/OMP protocol change. ContentCodec/portable AST, secure native storage, real provider/host/device acceptance and independent interoperability remain open. ADR-004/005/006/007/014/015/016/017/018/019/020 and R01–R07 in the PR plan carry the changes. The complete 54-unit implementation roadmap remains; passing this repair does not collapse the rest of the beta-to-RC work.

Current reconciliation: 2026-10-31, below. Earlier checkpoints remain historical where superseded. Proposed repairs require merge and CI; local evidence is not deployment acceptance.

Draft translation, exchange and authority checkpoint — 2026-10-31​

The new Draft implementations are worth keeping, with a targeted privacy repair and small ownership extractions. No editor, store, collaboration or history rewrite is justified.

The urgent defect is private annotation disclosure in both translation HTML proofs. The repair filters output and readiness through the existing public-note predicate, retaining all private authoring data. Translation review is not disclosure authority or scientific peer review. Private notes never become an authoring lock or an unrelated public-export prerequisite.

Translation commands move behind a synchronous application session port; Zustand/auth stays in a compatibility adapter. The current extension, history operation and Tiptap path locators remain explicitly experimental until a versioned editor-independent mapping is accepted. Standard export-report lifecycle moves into an application service while renderers and sidecars remain. Complete registry, streaming and target semantic validation are still follow-up work.

Reconcile the implemented single-node collaboration authority: current membership/ACTIVE account, per-message and 15 s idle checks, five-second authorization deadline, read-only downgrade and fail-closed disconnect/reconnect. Distinguish the new validation/exchange/self-assessment/core-review artifacts from publication evidence and exact-candidate RC approval. New core Drafts and a shared bounded validator do not silently change producer versions or supply independent conformance.

T01–T04 in the PR plan provide the immediate repair; T05/T06 retain the remaining freeze/acceptance obligations. Update ADR-001/004/005/007/008/009/011/012/017/019/023 as mapped in the register. No schema/container/DB migration or manuscript-editing restriction is introduced. The Studio companion must pass CI before these proposed repairs count as merged guarantees.

Security boundary reconciliation — 2026-10-30​

The new Studio security controls through 540f67a reinforce the target architecture and should be kept. The drift is an incomplete baseline/rollout contract, not a reason to replace identity, connectors or the PDF engine.

  • Admin bootstrap now requires an ACTIVE allow-listed account and explicitly verified matching OIDC/SAML email. Review older grants privately; test a second owner and trusted-IdP recovery. Missing claims must not lead to verification bypass.
  • Unsafe cookie writes require approved Origin. Native bearer and exact signed PKP-launch exceptions retain downstream authentication; invalid bearer must not fall back to a cookie. Document the 403 transition for older clients.
  • PDF imports have process-local admission limits (user/process: 1/2 active, 12/100 retained), retryable 429/503 errors and volatile job state. Workers receive allow-listed environment variables. These controls do not establish cluster quotas, durable recovery or OS/network isolation.
  • Repair the operator guides and add HTTP/provisioning regression coverage in both CI and readiness. The coordinated changes introduce no schema/container/DB migration and impose no restriction on manuscript editing.

The largest remaining security risk is operational: worker filesystem/resource isolation (#774), actual connection-time egress protection (#765) and candidate-specific deployment acceptance. They remain mandatory RC blockers; a green unit test cannot close them. Before freeze, record precise authority, Origin exceptions and PDF error/limit contracts in ADR-012/015/016/017/019. Before RC, aggregate deployed isolation, browser/native/PKP, recovery and other existing release gates for the same commit/artifact digest. See S01–S05 in the PR plan and the rollout runbook.

Audit baseline: 2026-09-19; Studio main @ eca2cf45762116e1c00a89b3397840ed89109511, OMI @ e2f421707f1154c0ed30cb7750e0689a837e9b0a, OJS plugin @ 6c0842fc1ce6c4a1ea75f7b27660af8011be303b, OMP plugin @ 4f7a2c0a2f29f68e3ad05ac73ca7652bc6410ac7.

This document is an implementation-architecture baseline for Open Manuscript Studio. It does not redefine normative OMI specifications and does not make a formal conformance claim.

Position​

The current Open Manuscript Studio codebase is a viable foundation for 1.0. A full rewrite is neither necessary nor desirable. The product already contains substantial working implementation: a rich Tiptap editor, study-level progressive mounting, immutable revision semantics, multi-format import and export, publication-build and artifact provenance, OJS/OMP review and writeback flows, account and institutional identity, cloud storage, Tauri and Android layers, and broad CI/test infrastructure.

The 1.0 problem is therefore not primarily missing functionality. The central problem is that several critical contracts are not yet aligned with the implementation that already works.

The original audit identified a 0.1/0.2 format mismatch. That specific mismatch is historical: Studio now uses OMI-SPEC-320@0.3.0 Draft with an executable vendored schema. The current source-attribution model introduces new lifecycle and export-fidelity obligations; #674 proposes targeted repairs. Stable schema/AST freeze and independent release evidence remain pending.

The recommended strategy is controlled refactoring:

  1. establish the contract or adapter in front of the current working implementation;
  2. lock existing behaviour with characterization, golden-fixture, and conformance tests;
  3. migrate call sites incrementally;
  4. remove the legacy path only after consumers have moved and rollback evidence exists.

How much of the implementation should remain?​

The architecture audit indicates that roughly 75–85% of the current implementation and behaviour can be retained.

The estimate is architectural rather than a formal line-of-code measurement:

  • about 45–55% belongs in KEEP/HARDEN;
  • about 25–35% can remain behind wrappers, ownership separation, or module relocation;
  • only about 5–10% appears to require genuine replacement or removal.

Refactoring effort may still be substantial because some consumers directly parse Tiptap JSON serialized into OMI block content and some application use cases currently mutate state directly through the Zustand store.

Explicitly preserve​

The 1.0 work should preserve and harden:

  • the Tiptap editor, semantic extensions, and study-based progressive mounting;
  • the current linear immutable revision and revert semantics;
  • the strong ZIP/container security controls already present in OMI container import;
  • the DOCX importer and the JATS, HTML, DOCX, PDF, and DTP exporters as implementation assets behind common contracts;
  • publication-build manifests, hashing, and provenance foundations;
  • the cloud-storage provider layer and SSRF protections;
  • the Tauri shell, Android distribution/update work, and platform build workflows;
  • the signed-launch, scope, and PKP object-authorization logic in the OJS/OMP integrations;
  • the server-side review allowlist projection and visibility serialization as the basis of anonymous review.

Replace or isolate narrowly​

The following areas should not be preserved in their audited form:

  • the empty Studio schema artifact;
  • native JSON open/save paths that rely on parsing and casting without exact schema/version validation;
  • native bearer-token storage in localStorage;
  • publication signatures injected into manuscript roots from local storage;
  • client-controlled authority for review-confidential integration access;
  • empty domain placeholder modules and demonstrably orphaned alpha workspace state;
  • the concrete MIDI import routing defect in which the parser exists but .mid and .midi fail the outer extension route.

Principal 1.0 architecture risks​

1. Portable format and data integrity​

The original OMI-SPEC-320 0.2 versus Studio 0.1 mismatch is resolved by current 0.3 Draft adoption; stable format and portable AST conformance remain blockers. Open/save/export must share one exact schema plus semantic-invariant validation path. A newer unsupported document must never be silently opened, edited, and overwritten. Unknown extension data also needs demonstrated lossless preservation.

2. Tiptap JSON as portable domain data​

The audited OmiBlock.content can hold serialized Tiptap JSON or legacy text. Citations, notes, cross-references, search, proofing, clipboard, and exporters know this representation in multiple places. That turns editor-runtime representation into a file-format dependency.

The solution is not replacing Tiptap. The solution is freezing the required OMI-SPEC-100 portable content grammar and introducing one versioned, loss-reporting Tiptap ↔ OMI codec.

3. State and use-case ownership​

The audited useStudioStore.ts mixes portable manuscript state, history, selection/focus, UI panels, persistence operations, and timers. Core use cases should be callable and testable without React or Zustand. Zustand should remain as a UI/session projection while mutations flow through an application facade.

4. Identity, credentials, and review security​

The two Prisma schemas partially duplicate account/session/identity concepts. The identity database should be formalized as the authority for account/authentication/membership, while the main database keeps a StudioPrincipal projection for workflow references.

Native credentials must move to platform secure storage. Double-anonymous review must remain a server-side security boundary and must cover asset metadata, filenames, EXIF/SVG content, and provenance as well as visible text. Confidential integration and AI access must be authorized by server-issued grants, never a client boolean.

5. External writeback and release evidence​

OJS/OMP remain authoritative for their own workflow state. A local Studio database and an external publishing system cannot be updated in one ACID transaction. Writeback therefore requires durable outbox semantics, idempotency keys, and receipts.

At the same time, a generally green CI run is not sufficient 1.0 evidence when some workflows are path-filtered. Release-candidate evidence must be aggregated on one exact commit without path filtering.

Required work before schema/API freeze​

The following items are architecture-freeze prerequisites:

  1. Freeze the portable content contract. OMI-SPEC-100 must identify the block/inline AST, stable object and anchor semantics, and extension points required by Studio 1.0. Golden round-trip fixtures must prove the Tiptap mapping.
  2. Align OMI-SPEC-320 and Studio. Use a released canonical schema, pinned by checksum, with generated wire types or codecs, validation on open/save/export, future-version quarantine, and unknown-extension round trip.
  3. Freeze the OMI-SPEC-330 container contract. Derive a normative manifest/path/checksum/security profile and cross-repository golden fixtures from the already working secure container implementation.
  4. Define history and identity wire boundaries. Internal revision-storage algorithms may evolve, but the portable/exported history boundary and the meaning of agent, contribution, and evidence data must not remain implicit.
  5. Introduce additive /api/v1. Shared Zod/OpenAPI DTOs, standard errors, idempotency, and optimistic concurrency should sit in front of shared handlers. Existing routes remain compatibility facades during migration.
  6. Freeze a common Publishing System Connector contract. OJS and OMP should share capability, launch, submission, file, review, writeback, and receipt vocabulary. Journal/monograph differences become profile extensions.
  7. Accept review and integration security boundaries. Anonymous projection is server-side; integration execution uses server-issued scoped grants.
  8. Enforce dependency direction. Core modules must not import React, Tiptap, Zustand, Tauri APIs, Prisma clients, or provider-specific DTOs.

What should not be rewritten now?​

The 1.0 programme should not replace Tiptap merely to obtain a cleaner conceptual model. It should not introduce a new history product or a CRDT requirement when the present immutable revision model already satisfies the current product semantics. It should not rewrite the ZIP parser, nor reimplement the DOCX/JATS/HTML/PDF/DTP processors, nor replace the OJS/OMP plugins with a generic plugin runtime.

Likewise, the entire backend route tree should not be renamed in one migration.

The following can be deferred from the stable 1.0 guarantee while preserving existing code as preview or experimental capability:

  • arbitrary executable third-party plugin runtime;
  • CRDT/branching collaboration;
  • a dedicated SharePoint provider beyond the existing storage direction;
  • stable iOS promotion until platform acceptance evidence is complete;
  • stable fidelity claims for every DTP format.

Phase A — Architecture freeze​

Accept or explicitly defer the ADRs, freeze stable/preview scope and compatibility policy, make OMI conformance CI mandatory, vendor the released schema, introduce shadow envelope validation and future-version quarantine, freeze content/container/API/connector candidate contracts, and enforce dependency boundaries.

This phase should not rewrite the working UI.

Phase B — Core refactoring​

Introduce the application facade and migrate create/open/save first. Narrow Zustand to UI/session projection. Put every portable-content consumer behind the codec before changing the stable writer. Place revision semantics behind a repository port. Make save atomic and recoverable. Formalize identity authority and move native tokens to secure storage.

Phase C — Interoperability hardening​

Wrap existing importers and exporters in shared contracts with diagnostics, fidelity, validation, and provenance. Separate artifact generation from delivery. Make publication profiles explicit versioned inputs. Require OJS and OMP to satisfy the same connector contract suite. Route writeback through an outbox. Make review projection and integration grants release-gated security behaviour.

Phase D — Platform hardening​

Implement common file, secure-storage, authentication, updater, font, and share ports across declared platforms. Use evidence to determine which platforms are stable versus preview. Add 10k, 120k, and 500k-word synthetic performance tests, bundle budgets, and accessibility journeys.

Phase E — Release Candidate​

One exact-commit RC aggregator runs format, data-integrity, import, publication, review, connector, identity/API, platform, security, accessibility, performance, and recovery gates without path filters. Private real-document acceptance remains local; only content-free results and synthetic minimal reproducers enter Git or CI.

Phase F — 1.0​

Promote the proven RC commit to 1.0.0 without functional change. Any code change after the RC evidence set requires a new RC. From 1.0 onward, file-format, API, and connector compatibility can change only under the published compatibility policy.

Final recommendation​

The highest-value work before 1.0 is not another wave of features. It is closing the format/content boundary, application-state ownership, security authority, and exact-RC evidence model.

The valuable implementation already exists. By adding explicit contracts around it, proving current behaviour, migrating incrementally, and deleting only obsolete paths after evidence exists, Studio can move from beta to a maintainable 1.0 without a disruptive rewrite.

Website publication supplement — 2026-09-23​

Supporting journals and presses without OJS/OMP is part of the 1.0 direction, not a weakening of scholarly editing. Studio-native review can supply a completed scientific round plus a separate editor acceptance bound to the exact revision/state digest. Website connectors only deliver artifacts; external publishing systems retain their own workflow authority.

Public-interest and popular-science publishing may proceed without review, with explicit disclosure. The language-independent seal is OMI · PEER REVIEW · VERIFIED or NOT VERIFIED, accompanied by localized explanation. It reports recorded workflow evidence, not scientific quality or truth; confidential reviewer data stays private.

Keep the current HTML renderer, publication build and connector foundations. Before freeze, settle the assurance contract and native-vs-external authority (ADR-021). Harden implementation under C15, with website delivery Preview until real database/receiver, recovery, privacy and accessibility tests pass. False assurance, reviewer leakage and missing required disclosure remain mandatory release blockers. Public verification portals, expanded review taxonomies and new workflow dashboards are not prerequisites for base 1.0.

2026-10-09 update​

The Studio's OMI-SPEC-320@0.2.0 runtime validator, vendored Draft schema and round-trip fixture now provide stronger implementation evidence than the original audit baseline. They do not freeze the format as stable or close the remaining future-version and independent-conformance gates.

The most urgent newly observed boundary was the research module insertion UI: it could flatten locally saved module records, including sensitive interview material, straight into portable manuscript content. Studio PR #630 contains that transfer while retaining section insertion. Keep discipline workspaces separate from OMI and follow with versioned project storage and an explicit, source-attributed and previewed contribution flow. These are targeted changes; the editor, module calculators, and publication pipeline do not need a rewrite.

External object source policy (ADR-023). File and clipboard visual objects now have a proposed per-object source confirmation gate in Studio PR #632, with automatic filename/part prefill. The current module excerpt PR #631 requires a selected excerpt and visible source. This is partial coverage: DOCX/PDF, PKP intake, reference-manager and other paths remain tracked in Studio issue #633. Do not claim universal enforcement: structured derivation linkage now exists under OMI-SPEC-225, but it is distinct from a citation occurrence and still needs lifecycle/export/privacy coverage.

Source attribution rollout. Studio PRs #634, #635, #636 and #637 extend the earlier visual/module source gate. The rule is implemented by ingress owner, preserving PKP review anonymity and canonical OMI open behavior. The source is presently visible text or the bibliography record itself; semantic object linkage and renderer conformance remain open in #633.

2026-10-10 drift checkpoint​

Current Studio main exposes four notable 1.0 gaps: an IDML hidden annotation projection leak, a DOCX source-font hint carried by transitional Tiptap JSON, native server selection coupled to Tauri/localStorage, and main container images labeled latest without the complete RC gate. Proposed, independent repairs are #654, #657, #656, and #655. They are open proposals until merged and verified. Preserve the working DOCX font fidelity path during codec migration; block any public artifact that leaks hidden provenance. A merged beta image change is still not an RC approval: portable AST round trips, native secure storage and exact-commit release aggregation remain mandatory before the 1.0 freeze.

2026-10-11 merged-state reconciliation and remaining gates​

Studio main now includes #654–#657: the previous IDML leak, readiness path filtering, native server-selection placement and uncharacterized source-font decoding have targeted repairs. Keep these tested implementations.

The next verified risk is cross-format annotation disclosure. #662 proposes a shared public manuscript projection for eight formats, with synthetic fixtures retaining public notes and excluding private bodies/anchors without modifying the source document. It remains subject to CI and merge. EPUB note-body fidelity and PDF binary/editorial acceptance are explicitly separate.

The remaining pre-RC work is still substantial: native secure storage and origin-scoped cache/account ownership, versioned ContentCodec/portable AST decisions and E02 exact-commit artifact/evidence aggregation. The merged containment work improves the beta but does not establish a stable schema/API freeze or 1.0 readiness. No full rewrite is justified by these findings.

2026-10-12 current implementation checkpoint​

This checkpoint supersedes earlier dated implementation-state descriptions; it does not retrospectively change their audit evidence or declare architecture freeze. Studio main was verified at 6242dbae4e6e7be3f9bff9470f64647554f363d8; OMI at bc932272d0cbe18b8e45f758fbdb0d6076344bbf.

  • Studio now writes/validates OMI-SPEC-320@0.3.0 Draft with schema and omi.version, and retains the separately versioned container/history layers. Do not add a redundant schemaVersion field or promise pre-stable 0.2 migration. Stable compatibility obligations still begin with the first frozen stable generation.
  • OMI-SPEC-225@0.1.0 defines sourceAttributions: independently identified prov:wasDerivedFrom links to bibliographic records with explicit machine-inferred / author-declared / author-confirmed status. These are distinct from citation occurrences, technical import provenance, rights and authorship evidence. The model exists; universal ingress/export coverage does not.
  • Studio PR #674 proposes atomic source-relationship cleanup on block-subtree deletion, retaining unrelated relationships, bibliography records and prior revision snapshots. It also proposes explicit HTML/JATS source-attribution-not-preserved warnings. Until merged and verified, these are repairs under review, not main guarantees. The warnings disclose loss; they do not establish a lossless mapping or generate a citation/licence claim.
  • Mandatory regression evidence for this repair is actual insertion → checkpoint → subtree deletion → checkpoint → OMI save/reopen → historical revert, plus HTML/JATS per-relationship loss diagnostics and non-mutation. Fixtures must be synthetic/public. Broader deletion/merge paths, additional exporters, anonymous projection and OMI-container source-graph round trips remain separately assessed gates.
  • Preserve the working editor, snapshot history and renderers. Portable AST/ContentCodec, immutable schema provenance, unknown-field/future-version handling, secure native storage and exact-commit RC evidence remain prerequisites; 0.3 adoption is not 1.0 readiness.

2026-10-13 — ODT Preview trust and fidelity checkpoint​

The earlier #674 source-lifecycle repair is merged at 5825c15a1bc3860a3371eaa1c20111bfe4b74ea5; its prior proposal checkpoint is historical.

ODT import/export entered Studio main in #675 (commit 5417e19cb26550508b66c780107eea87382b5f2f). It adds an interoperability format, not a new OMI schema, account authority or durable working-state format. ODT is Preview, pending independent ODF/office fidelity and real-document acceptance; it must not be promoted merely because the ZIP can be opened.

Studio #678 proposes targeted containment of two reproduced gaps. The import reader had trusted declared ZIP sizes and fully materialized DEFLATE output before checking it. The exporter omitted sourceAttributions without loss reporting. Treat #678 as a proposal until merge and verification, not a main guarantee.

The proposed boundary retains the 200 MiB input-package limit and enforces 25 MiB per expanded part, 100 MiB aggregate expanded data, and 4096 entries. Actual streamed bytes trigger cancellation before complete output materialization. Local/central headers, bounds, CRC32 and duplicate/unsafe/overlapping entries are checked; encryption, symlinks, ZIP64 and multidisk are explicitly unsupported. Successful repeated reads share the cached budget. CRC is integrity checking, not authenticity.

ODT uses the shared source-attribution-not-preserved diagnostic per unsupported relationship, alongside existing visible delivery warnings. A source label is not a semantic derivation link; no citation, licence or authorship claim is fabricated. Portable OMI retains the authoritative graph/history.

AreaDecision / release gate
ADR-008 / ADR-017 — importHARDEN the existing ODT adapter with real output-byte limits and ZIP integrity. Forged STORE/DEFLATE lengths, aggregate/entry quotas, CRC/header corruption, duplicate paths, truncation and ordinary data-descriptor compatibility are mandatory synthetic regression cases. Preview does not waive security gates.
ADR-008 / ADR-009 / ADR-023 — exportReport loss of every unsupported source relationship, with stable code and object target; preserve original portable state and keep diagnostic messages content-free. A later target mapping needs independent export/reimport proof.
ADR-019 — capabilityODT remains Preview with documented supported subset and losses. ODF schema/independent office acceptance and real-document fidelity gate Stable promotion.
PR planMerge/review #678 as the bounded incremental repair after #675; retain the working import projection and delivery paths. Then prove independent ODF validity and cross-office fidelity before any Stable claim. No beta-format migration or renderer rewrite is justified.

This checkpoint supplements the earlier format matrix and source-graph gates. The original schema/API/ContentCodec/native security/exact-commit RC prerequisites remain in force.

Architecture repair checkpoint — 2026-10-15: workspace mode and passkeys​

ADR-001/002/005: simple writing is a Studio workspace preference keyed by manuscript ID, persisted locally outside portable manuscript state and revision snapshots. New documents do not write editorMode; the Studio vendored 0.3 schema must match the OMI source artifact. Existing unknown fields remain preserved under the general round-trip policy; this repair does not silently delete an imported extension or reinterpret it as authorization. Moving a file to another installation does not transfer local writing-mode preferences. The simple visual editor remains available.

ADR-012/015/016/017: passkeys authenticate the existing local account; they do not establish scholarly identity or publication evidence. /api/v1/auth/passkeys is the client contract. /api/auth/passkeys remains a compatibility facade over the same handlers and rate limits. HTTP routes own transport/status/cookies, the passkey service owns WebAuthn use cases, and the repository adapter owns identity-database access. Challenges remain expiring and single-use; verification still binds RP, web origin, user handle and user verification. Credential material never enters OMI.

Support is web-origin only: platform-authenticator availability is checked behind a capability adapter. Tauri origins do not advertise this flow. Native promotion requires a verified browser/native handoff and contract tests; do not widen accepted origins to make native UI appear supported. PostgreSQL migration, real authenticator registration/login/revocation, negative origin/RP/signature cases and native handoff evidence remain release requirements, not claims established by mocked route tests.

Validation for the repair includes per-document workspace isolation, unchanged portable bytes/history, schema parity, native-origin capability rejection and v1/legacy route equivalence with replay rejection. This checkpoint records a proposed repair pending merge and CI, not a completed 1.0 freeze.

External integration stabilization — 2026-10-18​

Keep the functioning EuroOffice embedding, Transkribus client, model picker, source attribution and OMI editor. The required changes are bounded: safe callback persistence, validated streamed DOCX reception, v1 adapters/shared transport and a durable paid-execution intent. Ordinary DOCX remains outside OMI; no format rewrite is needed.

Force saves must not be invalidated after the first save. They become recoverable snapshots; final close commits the canonical revision with a precondition and prior-head backup. A late callback cannot replace newer canonical content. Owners can download recovery snapshots without implicitly restoring them.

Recognition submissions persist the exact image digest/model/connection/grant and request identity before spending credits. Accepted retries return the existing job; ambiguous submissions stay unknown and are not automatically reissued. This reduces duplicate spending but cannot guarantee exactly-once billing at an external provider. Confidential review export is not authorized by this uploaded-image capability.

Before deployment, drain existing EuroOffice sessions, back up the database and apply the additive migration; old unbound sessions must reopen. Merge requires synthetic regression and PostgreSQL migration/concurrency evidence. Real-provider and native-device acceptance remains necessary before Stable promotion. ADR-024/025 record these boundaries. This checkpoint describes proposed repairs pending merge/CI, not a completed release freeze.

2026-10-21 — Federated discovery repair checkpoint (ADR-026)​

The new server directory is an operational discovery authority, separate from account, manuscript, review and publication authorities. DNS proves domain control only. Heartbeat online denotes a recent authenticated heartbeat, not service health, compatibility, capacity or trustworthiness. Browser selection opens the chosen origin; native selection retains the existing origin/token boundary. Discovery never proxies application traffic or transfers accounts/documents automatically.

The repair introduces /api/v1/server-registry; the old /api/server-registry is a shared-handler facade. The route delegates to application service, repository and GitHub audit adapters. Registration needs a private claimant capability distinct from public DNS data, expires in 24 hours, and uses database CAS for PENDING → VERIFYING → ACTIVE. Unknown audit outcomes stay quarantined without automatic issue resubmission. ACTIVE credentials survive the additive migration; old PENDING entries must re-register. Rotation/revocation/heartbeat updates compare current state and token atomically.

This is a proposed repair checkpoint, not a completed 1.0 freeze. Mandatory evidence includes synthetic negative/dual-mount tests, actual PostgreSQL migration/concurrency/lifecycle tests, plus controlled external DNS/GitHub acceptance and operational recovery. Deploy migration → API → clients/CLI/website; keep discovery Preview until acceptance evidence exists. Origin-scoped account/cache ownership and native SecureStorage remain separate open work.

2026-10-22 — Document import origin clarification (ADR-023)​

This checkpoint supersedes the requirement to anchor whole-manuscript own-work declarations to imported content. OMI-SPEC-225@0.2.0 permits own-work imports without bibliographic attribution. An explicit own-work choice for DOCX, ODT or PDF stores the version-1 https://openmanuscript.org/extensions/manuscript-import-origin/1 document extension: sourceFormat, importedAt, and assertion: author-declared:self-authored-outside-omi. It is self-attestation, separate from a citation and verified authorship. No filename, path, URL, account identity, section ID, block ID or anchor is stored. Other import paths retain their existing per-object policy.

Every document part remains editable, replaceable, movable and deletable. The declaration is document metadata and creates no content relationship; replacing all imported content does not produce dangling anchors. It records a historical import declaration, not authorship of later edits. Working OMI, history snapshots and core-snapshot interchange preserve it via the existing namespaced extension contract, without a schema-version change. Public publication projections omit the extension without mutating the working document; anonymous server review projections remain allow-listed and exclude it.

Validation for the repair: DOCX/ODT/PDF own-work intake, deletion/replacement of imported content and titles, save/reopen, history-free core-snapshot round-trip, revision snapshot retention, public export omission and anonymous review filtering. Keep the existing editor, import parsers and revision engine. No database migration or legacy-marker conversion is required. Deployment status remains proposed until the coordinated Studio repair is merged and CI passes.

2026-10-25 — Publisher profile import repair (ADR-009)​

External file, URL and catalog profiles must pass a shared nested contract validator before activation. Validate required rule groups, enums, finite bounded layout values, citation style identifiers, protection shape, attached style structure and JSON complexity; reject prototype keys. Invalid input must not enter storage or change the active profile. The import boundary records SHA-256 over the exact received bytes for all three paths; catalog imports additionally compare the independently supplied catalog digest. A digest identifies content, not publisher identity. Imported protection metadata remains an unverified read-only declaration, never an authorization credential.

Generic scholarly and humanities profiles with independent note/bibliography defaults advance to 0.2.0. Retain explicit 0.1.0 definitions; resolve known versions exactly and reject unavailable versions instead of silently selecting newer behavior. Existing publication-build manifests already hash the resolved profile; retain that digest independently from local download provenance. Changes to render-affecting profile defaults require a new profile version. No application release bump or OMI-schema migration is needed for this repair.

Release evidence: negative nested-input fixtures, checksum mismatch and malformed UTF-8 rejection, deterministic byte hashes, old/new profile resolution, existing publication-build and renderer regression suites. The validator remains an application contract for the alpha profile model, not a newly frozen normative OMI schema. Further limits for streamed network reads and independently authenticated publisher signatures remain hardening work; checksum comparison alone does not establish trust. This checkpoint is proposed until the coordinated Studio PR is merged and CI passes.