docs: maintain durable product-technical gap baseline

[seonghobae]

Current review admission — 2026-10-01

• Exact head remains 301f843c01f301088234c27f758c2a074cfa9aed; protected base remains develop@eb9757f8649aaad026a9865508d9aad50c1a7a4f.
• State: Ready for review / Proposed / merge HOLD, mechanically mergeable.
• One ordinary Draft → Ready transition followed exact-head Foundation/SAST GREEN, 91 COMMENTED reviews, zero unresolved threads across 65 review threads, and zero qualifying approvals.
• Ready is review admission only. Dependency Review HTTP 403, fresh CodeQL admission, and a qualifying independent approval remain mandatory merge gates.
• No source mutation, empty/no-op commit, manual rerun, Draft/Ready cycle, self-approval, bypass, merge, auto-merge, release, or closure was used.

---

Scope

Canonical single-writer lane for docs/product-technical-gap-baseline.md. The baseline is a durable commercialization map, not merge authorization: it records protected/released truth and explicit Planned/Active gaps without promoting Draft implementation to shipped capability. Volatile workflow IDs, queue states, mergeability and reviews remain live GitHub truth.

Current authority — 2026-09-30

• Protected truth remains develop@eb9757f8649aaad026a9865508d9aad50c1a7a4f.
• This canonical baseline branch is exact 301f843c01f301088234c27f758c2a074cfa9aed, open · Ready / Proposed. Ready admits review but is not protected truth, approval, merge authorization, or release evidence.
• docs/product-technical-gap-baseline.md is now a required manifest-sealed Foundation artifact with linked PRD/TRD/UML/ERD/Context Map views, explicit Gap/Action/Status, and 17 exact-head coordinates covering every Active PR gap plus the causal composition path.
• API-01 remains Planned. feat(composition): prevalidate complete ASGI responses before start #446 is Draft exact 3a20660761bc27de74a89c4027d9a89cf74776eb; ci: execute discovered service Python compatibility #447 is Ready exact c19e22d93c4b9e3d6a86f66b69f7e581c10cb249. Neither creates protected deployable composition, immutable released external authority, or buyer-path SLO evidence.
• RED proved the baseline was absent from the required Foundation inventory and lacked the evidence index, explicit Action/Status fields, and reproducible Active-PR coordinates. GREEN is npm run validate: 58/58 Node tests plus Python manifest/repository validation and git diff --check.
• Independent review findings were repaired: external ACL wording is release-gated, every Active PR gap has a 40-hex exact-head coordinate, unsupported GOV-01 remains Planned, and the regression enforces the mapping.
• Exact-head hosted evidence for 301f843c01f301088234c27f758c2a074cfa9aed: Foundation CI run 36796830293 and SAST run 36796830323 succeeded; Security run 36796830339 failed closed because Dependency Review job 110162113278 received HTTP 403 with curl_exit=0 for exact base/head; CodeQL run 36796830370 was skipped and is not acceptance evidence. Qualifying independent approvals: 0; unresolved review threads: 0. No predecessor evidence transfers.

Durable gap states

ASM-01, INT-01, and VAL-01 keep their existing owner paths. API-01 remains Planned through #432, Proposed architecture #433, and the Draft product-composition stack through #446. Shared-edge configuration, structural admission, ASGI correctness, or process-local evidence integrity alone does not close the buyer-visible Gateway gap.

The durable owner dependency for API-01 has materially changed and must be reflected in baseline source: repository-level service/runtime execution now has a concrete canonical implementation lane (#447), not merely an unresolved #260 concept. The causal path is #259 plus truthful #305/#64 metadata integration → #447 Foundation discovery/runtime acceptance → #261 installed-artifact acceptance + #311 PostgreSQL discovery/execution → composition ordinary-forward adoption and one unchanged exact-tree package/PostgreSQL acceptance → #433 independent architecture admission → released Keyverse/Orgmetra/owner API authority → deployable authenticated ASGI host, owner dispatch, cleanup/reload/recovery → realistic [shared edge if deployed] -> composition -> owner HTTP -> PostgreSQL k6/E2E with applicable p95 ≤20 ms → immutable release evidence.

This owner-path currentization does not change API-01 from Planned and does not make #447 or #446 shipped capability.

API-01 current interpretation

Protected Architecture/API truth advertises a buyer-facing Orgmetra Gateway while protected executable truth still lacks a supported deployable product-composition boundary.

The durable ownership model remains optional released shared-edge transport plus a separately deployable Orgmetra product-composition application. Composition owns product route admission and request-scoped projection only, not HR truth, Person binding, purpose authorization, owner idempotency/replay/concurrency/error semantics, or scientific truth.

The #434→#446 lineage contributes bounded admission, durable configuration/currentness, routing/request-body and response-lifecycle invariants without changing the Planned gap state. These mechanisms remain Draft source evidence rather than durable release/configuration/deployment authority.

Canonical verification is now explicitly Foundation-owned: #447 executes discovered services from service/package metadata and mandatory owned closure; #261 proves installable wheel/sdist behavior; #311 proves owner-neutral PostgreSQL contracts. The composition stack must consume those capabilities after normal integration and prove one unchanged exact candidate rather than infer execution from another service or use a feature-local workflow/source-tree packaging shortcut.

Shipment still requires released Keyverse/ACL and owner API/operation evidence, durable non-reassignable generation/config/deployment activation authority, canonical Foundation execution of the service, a deployable HTTP composition host, activation/rollback re-admission, fault/security/recovery evidence, supported Kubernetes packaging, and realistic asynchronous [shared edge if deployed] -> composition -> owner HTTP -> PostgreSQL k6/E2E with applicable p95 <=20 ms.

Baseline write rule

Planned/Active work belongs in baseline source when it changes durable buyer/scientific gap state or owner dependency and is explicitly labeled. A Draft PR, local result, ADR or metadata update never becomes Shipped truth merely by being recorded. Conversely, commit churn that does not change durable state stays in owner PR/issue authority rather than repeatedly rewriting this baseline.

This branch remains the only writer for docs/product-technical-gap-baseline.md. Future source currentization must preserve truth-state distinctions, exact owner boundaries, causal stack order, standards-profile scope, canonical identity grammar, process-local evidence integrity versus durable provenance, and fail-closed external-contract semantics.

2026-10-01 single-writer repair

• Re-fetched feat(outbox): integrate governed delivery receipt into Foundation #448 exact head 8f3cc0d7fd217cbc7f3b70431286110bbce71f47: it was Draft, with Foundation CI/SAST terminal-success, Security Scan/CodeQL terminal-failure, zero qualifying approvals and zero unresolved threads.
• RED: the baseline mislabeled feat(outbox): integrate governed delivery receipt into Foundation #448 as Ready API-01 evidence and had no governed outbox-receipt Gap mapping.
• GREEN: exact head a7d3f3b74f6ca28bd2dd5ad81e60fad376b6019a records ORGMETRA-OUTBOX-RECEIPT-01, adds a buyer/audit consequence and causal acceptance action, and expands the regression parser to support multiword semantic Gap identifiers.
• Canonical validation: npm run validate passed Python repository/manifest validation and 58/58 Node tests; git diff --check passed. manifest.json seals the changed baseline and regression test.
• feat(outbox): integrate governed delivery receipt into Foundation #448 separately removed its duplicate baseline file while preserving its outbox package, ADR, traceability, workflow and test delta. This PR remains the sole writer for docs/product-technical-gap-baseline.md.
• 301f843c01f301088234c27f758c2a074cfa9aed refreshes the canonical row and regression fixture to feat(outbox): integrate governed delivery receipt into Foundation #448 current head 6a06062e9d4f6e07bf4780a60384f264f7983353; npm run validate remains GREEN at 58/58 Node tests plus Python repository/manifest validation.
• Merge/release authorization is unchanged: this PR is Ready only for review admission and remains merge HOLD pending terminal exact-head Security/CodeQL evidence and a qualifying independent approval.

[coderabbitai]

Note

Reviews paused

It looks like this branch is under active development. To avoid overwhelming you with review comments due to an influx of new commits, CodeRabbit has automatically paused this review. You can configure this behavior by changing the reviews.auto_review.auto_pause_after_reviewed_commits setting.

Use the following commands to manage reviews:

• @coderabbitai resume to resume automatic reviews.
• @coderabbitai review to trigger a single review.

Use the checkboxes below for quick actions:

• ▶️ Resume reviews
• 🔍 Trigger review

📝 Walkthrough

Walkthrough

제품·기술 격차 기준을 2026-09-16 기준의 durable commercialization baseline으로 재작성했다. 보호된 develop 상태, 상업화 격차, ADR·추적성 상태, 모델 라우팅 및 소유권 정책 검증을 갱신했다.

Changes

제품·기술 격차 기준

Layer / File(s)	Summary
상업화 기준 및 릴리스 게이트 docs/product-technical-gap-baseline.md, docs/doctoring/product-gap-baseline-references.md	제품 테제, truth-state 계약, 도메인 맵, 14개 상업화 격차, 데이터·과학·AI 불변식, 보안·개인정보 기준, 연구 근거와 10개 릴리스 게이트를 추가했다.
보호된 develop 상태 및 추적성 CHANGELOG.md, README.md, docs/TRACEABILITY.md, docs/adr/*, docs/traceability/*	여러 기능과 ADR·추적성 항목의 상태를 active PR 또는 제안 상태에서 보호된 develop 구현·승인 상태로 변경했다. Naruon 통합은 calendar intent 경계로 재정의했다.
정책 경계 및 검증 .gitignore, AGENTS.md, CLAUDE.md, package.json, tests/llm-routing-policy.test.mjs	.codegraph/ 제외 규칙을 추가했다. 모델 기반 Actions의 orchestrator/free 라우팅과 ConceptWeave·semantic-data-portal·contextual-orchestrator의 소유권 경계를 문서화하고 테스트했다.
무결성 메타데이터 갱신 manifest.json	변경된 문서와 package.json의 해시, 바이트 수와 라인 수를 갱신했다.

Priority: ⬇️ Low

Estimated code review effort: 2 (Simple) | ~15 minutes

Change: Other

Merge Risk: 🔵 Low · up to ccfbf

The baseline can still permit inconsistent Talent denominator interpretation and make supporting research evidence harder to reproduce, but these are bounded documentation risks rather than immediate runtime failures.

🚥 Pre-merge checks | ✅ 4 | ❌ 1

❌ Failed checks (1 warning)

Check name	Status	Explanation	Resolution
Docstring Coverage	⚠️ Warning	Docstring coverage is 0.00% which is insufficient. The required threshold is 80.00%. Docstring coverage is scoped to functions touched by this diff. Analyzed 1 functions across 1 files. (1 skipped: 1 …	Write docstrings for the functions missing them to satisfy the coverage threshold.

✅ Passed checks (4 passed)

Check name	Status	Explanation
Linked Issues check	✅ Passed	PR 설명과 문서에 #63, #64, #65, #141, #163, #165, #235, #248, #398~#401, #772 등의 관련 이슈와 PR이 명시되어 있습니다.
Out of Scope Changes check	✅ Passed	변경 사항은 baseline 문서화, 저장소 정책, 추적성 상태, ADR 상태, manifest 갱신, 관련 검증 테스트 범위에 있습니다. PR 목적과 무관한 기능 구현은 확인되지 않습니다.
Description Check	✅ Passed	Check skipped - CodeRabbit’s high-level summary is enabled.
Title check	✅ Passed	PR 제목은 durable product-technical gap baseline 유지라는 변경의 핵심 목적을 정확히 설명하며, 간결하고 구체적입니다.

Full details: Docstring Coverage

Explanation

Docstring coverage is 0.00% which is insufficient. The required threshold is 80.00%. Docstring coverage is scoped to functions touched by this diff. Analyzed 1 functions across 1 files. (1 skipped: 1 unsupported.)

• Fix all pre-merge checks with AI

✨ Finishing Touches

🧪 Generate unit tests (beta)

• Create PR with unit tests
• Commit unit tests in branch docs/product-technical-gap-baseline

---

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

• X
• Mastodon
• Reddit
• LinkedIn

_{Comment @coderabbitai help to get the list of available commands.}

[devin-ai-integration]

Note

This report is out of date. Scroll down for Devin Review's latest report on this PR.

✅ Devin Review: No Issues Found

Devin Review analyzed this PR and found no bugs or issues to report.

[seonghobae]

@opencode-agent Please review the current unchanged head against protected develop. Local exact-head verification: all owned package suites pass at 100% statement/branch coverage.

[chatgpt-codex-connector]

You have reached your Codex usage limits for code reviews. You can see your limits in the Codex usage dashboard.
To continue using code reviews, you can upgrade your account or add credits to your account and enable them for code reviews in your settings.

[chatgpt-codex-connector]

You have reached your Codex usage limits for code reviews. You can see your limits in the Codex usage dashboard.
To continue using code reviews, you can upgrade your account or add credits to your account and enable them for code reviews in your settings.

[seonghobae]

Baseline handoff only; no source rewrite requested from this lane. #437 exact 27b82cc74a60838d57d31aa0a2ed3a131b243b95 now adds durable append-only restart re-admission evidence through migration 0022. This strengthens the Draft implementation but does not change the durable buyer-visible state: API-01 remains Planned because there is still no protected deployable composition application, canonical exact-head execution, real immutable external-release positive path, or buyer-path p95 evidence.

When baseline state eventually changes, preserve the distinction between historical activation evidence and fresh recovery-attestation evidence; do not treat process-local recovery return values or Draft migration existence as Shipped truth.

[seonghobae]

Gap-baseline handoff: #437 moved to exact 03eb4b0861a4f1de07c9b66c20f71ab28d378559 and repaired 0022 schema-publication atomicity. This remains Draft implementation hardening only; no protected deployable buyer path, immutable external production evidence, exact-head canonical execution, or release exists. API-01 should therefore remain Planned and docs/product-technical-gap-baseline.md should not be promoted to Shipped/Active from this delta.

[seonghobae]

Gap-baseline handoff only; no source write requested here. Draft composition hardening advanced #436 to 6ebe5ec... and #437 to c8886e6..., closing unguarded schema-publication windows in migrations 0018 and 0019. Buyer-visible truth has not changed: no protected deployable composition path, released external authority evidence, exact-head canonical execution, full-path p95 acceptance, or immutable Orgmetra release. Keep API-01 Planned and do not promote this Draft movement to Shipped/Active in docs/product-technical-gap-baseline.md.

[seonghobae]

Baseline handoff only; no source write requested from this lane. #437 exact 035826c9ddb1534290d8ed7957da3c5a45b80da4 now includes database-level activation/recovery deployment-row serialization through migration 0023. This is Draft authority hardening and does not change buyer-visible gap state: API-01 remains Planned until protected deployable composition, immutable external authority, canonical execution and end-to-end acceptance exist. Keep docs/product-technical-gap-baseline.md unchanged for this commit churn.

[seonghobae]

Gap-baseline handoff only; no baseline source write requested. Draft #437 is now exact 84d347b5257bbeb876cf7407a4eefd1f25257524 (110 commits / 37 files), with migration 0024 securing trigger-function provenance against migration-session search_path masking and current recovery/serialization PostgreSQL contracts running on that final schema. This does not change durable buyer-visible state: API-01 remains Planned because protected truth still lacks a deployable supported composition host, canonical exact-head execution, real immutable Keyverse/Orgmetra/owner evidence, release provenance, and buyer-path performance acceptance. Preserve the current baseline write rule rather than enumerate Draft implementation churn as shipped capability.

[seonghobae]

API-01 handoff only; no baseline-source write requested. Current Draft composition authority is #436 5fd0087179f2e6f23f2bb3853ad1de397fc53b0e plus #437 468dab6d8be4cdd8467b5093434e4756ce1ce9b0 (111 ahead / 0 behind parent). This run hardened 0018 generation-authority migration provenance against caller-controlled search_path and restacked the activation/recovery child non-force. It did not create a protected deployable HTTP composition path, immutable positive external releases, canonical exact-head PostgreSQL/coverage evidence, buyer-path p95 evidence, or an Orgmetra release. Therefore API-01 remains Planned and docs/product-technical-gap-baseline.md should remain unchanged for this commit-level hardening. #433 also remains Proposed/Draft and its source still needs currentization before architecture admission.

[seonghobae]

Baseline-owner handoff only; no baseline file edit from this lane. #437 exact 960a5fbfcb98075862cdbee8ee094a9b8d7dcc55 closes an ambiguous post-commit recovery-reporting defect by leaving PostgreSQL as commit-time freshness authority and removing a fallible local clock recheck after durable recovery attestation commit. API-01 should remain Planned: protected exact-head execution, immutable real external releases, deployable HTTP composition, and full buyer-path p95 evidence are still absent.

[seonghobae]

API-01 durable gap state remains Planned; no docs/product-technical-gap-baseline.md source write is justified by this Draft-only increment.

Current downstream composition leaf is Draft #446 exact 83f329279dc9e89a75df05c7b76a671cd1cfbd5c, stacked directly on #444 exact e1af671deeeec682dffdf809cb8579dcc4d439da. #446 adds a complete non-streaming response owner above the existing send/event primitive: response-start and terminal-body are prevalidated before the first send; 1xx is excluded from final-response authority; explicit HEAD suppression and 204/205/304 no-content semantics are preserved; explicit Content-Length is now fail-closed against RFC 9110 framing semantics (204 forbids it, 205 can describe only zero content, HEAD/304 and ordinary responses must match the owned representation length, duplicate/invalid values are rejected before transport).

This improves Draft HTTP correctness but does not create a protected deployable composition application, immutable released Keyverse/Orgmetra/owner authority, canonical exact-tree package/PostgreSQL acceptance, authenticated owner dispatch, cleanup/recovery, immutable release, or full buyer-path p95 evidence. Keep the baseline source state Planned and preserve #100 single-writer ownership; volatile exact heads remain live GitHub authority.

[seonghobae]

Exact-head currentization evidence for d883a1d58587df933a973665d96c109fb4f3b749.

• Canonical single-writer branch preserved; parent is aeacd8deda19f878d802c70b77c65912fb4a6e7e; ref advanced by an ordinary non-force commit.
• RED: the gap baseline was not a required Foundation artifact and did not expose the mandated PRD/TRD/UML/ERD/Context Map index, explicit Action/Status columns, or reproducible exact-head evidence for Active PR gaps.
• GREEN: npm run validate passed Python manifest/repository validation, Foundation validation, and 58/58 Node tests. git diff --check passed.
• The baseline is manifest-sealed and the regression requires every Active PR gap to map to at least one repository/PR plus a 40-hex exact head.
• Independent review findings were repaired and the re-review reported no Important findings; its final Minor scope wording finding was also corrected before the final GREEN run.
• API-01 remains Planned. Draft/Ready PR coordinates are evidence only and are not promoted to protected/released truth.

This PR remains Draft. Fresh hosted exact-head Checks and independent approval remain merge gates; no manual rerun, self-approval, admin bypass, Force Push, or destructive rebase was used.

[seonghobae]

Review-admission transition for unchanged exact head d883a1d58587df933a973665d96c109fb4f3b749.

Source readiness was re-established before transition: PR remains mergeable, unresolved review threads are 0, local exact-head Foundation/manifest validation is GREEN with 58/58 Node tests, and the independent re-review findings were repaired. The body now states the correct boundary: Ready is review admission; terminal exact-head hosted Checks and qualifying independent approval are merge gates, not Ready prerequisites.

This transition does not claim protected/released truth, transfer predecessor evidence, approve, merge, or authorize release. No empty push, manual rerun, self-approval, review dismissal, Force Push, destructive rebase, or administrator bypass was used.

[seonghobae]

Exact-head admission correction — d883a1d58587df933a973665d96c109fb4f3b749

Ready is review admission only. Fresh audit against base eb9757f8649aaad026a9865508d9aad50c1a7a4f found:

• latest terminal workflow blockers: Foundation CI 36289618026=cancelled, CodeQL PR 36289617980=cancelled, Security Scan 36289618036=failure

This PR is moved to Draft/Proposed until the causal owner repair is present on a successor exact head and re-audited. Queued/pending work is neither an additional blocker nor passing evidence. No Close, force push, destructive rebase, manual rerun, synthetic status/approval, merge, auto-merge, or bypass was performed.

[seonghobae]

Draft admission correction — 2026-10-01

Exact head 301f843c01f301088234c27f758c2a074cfa9aed is mechanically mergeable and all 65 review threads are resolved, but Security run 36796830339 failed when dependency-review received HTTP 403 after exact checkout identity succeeded. CodeQL is skipped and qualifying APPROVED reviews remain 0.

Move back to Draft / Proposed pending repository configuration owner #449, fresh exact-head Security evidence, and independent approval. No result is rerun, synthesized, weakened, or transferred.