litclaude-ai 0.3.35 → 0.3.37

package/CHANGELOG.md

@@ -1,5 +1,63 @@
 # Changelog
 
+## Unreleased — `lit-comprehend` explainer artifacts
+
+- Add the `lit-comprehend` skill, the `/litclaude:lit-comprehend` command, and the
+  `lit-comprehend` / `comprehend` / `lit comprehend` / `$lit-comprehend` /
+  `$comprehend` hook routes. The trigger surface is deliberately wider than the
+  skill id — bare `comprehend` is what users type, the same reasoning that keeps
+  bare `recap` routed to `lit-recap`.
+- Gate execution instead of narrowing activation: being invoked is not permission
+  to build. When the invocation names the file or commit set directly (a path, a
+  git range, a branch, a PR) the skill proceeds; otherwise it proposes the scope,
+  its countable size, what it excluded, and the rough cost, offers a one-sentence
+  answer when the request deserves one, and waits. The proposal is derived from
+  `git status` / `git diff --stat` / durable state, never by reading the tree — a
+  gate that costs as much as the thing it gates is not a gate.
+- Fix a routing defect found while wiring this: `lit comprehend` and
+  `litwork comprehend` matched the bare `lit` / `litwork` tokens first and
+  activated lit-loop, injecting the wrong skill body. The two-token phrase now
+  resolves ahead of the leading-bare check, with a regression test.
+- Where `lit-recap` answers *what happened*, `lit-comprehend` builds the artifact
+  a reader needs to reason about the work: it anchors on what they already knew
+  and explains only the delta, walks the change in conceptual rather than file
+  order, ships an interactive micro-world, discloses what is **not** verified,
+  and closes with a quiz framed as a speed regulator rather than a grade.
+- Bundle `assets/explainer-scaffold.html` so the CSS, diagram components, and
+  quiz engine are written once instead of per invocation, plus
+  `references/artifact-template.md` and `references/micro-worlds.md`.
+- Add `scripts/verify-explainer.mjs` as a real verification gate. It fails a
+  phantom code quote, a quote attributed to a missing file, an external resource
+  reference, a missing canonical section, an artifact written inside the
+  worktree, a quiz option without feedback or with a positional tell, collapsed
+  code blocks, and ASCII-art diagrams — the defect classes that leave a reader
+  confidently wrong rather than merely uninformed.
+- Pin the whole `skills/lit-comprehend` tree in the canonical resource manifest so
+  the installer, install-time integrity gate, `doctor`, and package payload
+  guard all cover the bundled scaffold and verifier.
+
+## 0.3.37 - 2026-07-29 — canonical frontend corpus and workflow families
+
+- Ship the exact canonical frontend library with byte-pinned manifests, legal
+  companions, independent commitments, bounded reads, and source-to-tar payload
+  verification while retaining the independent normalized design dataset.
+- Add Claude-native `autoresearch`, `autoconference`, and `wikify` skills and
+  command routes for their supported modes, including an explicit blocked state
+  when root multi-agent capability is unavailable.
+- Bind the three workflow-family adapters and their complete source closures to
+  canonical manifests, immutable commitments, installed integrity checks, and
+  package-payload verification.
+
+## 0.3.36 - 2026-07-28 — G8 UI/UX capability and evidence hardening
+
+- Promote the beta Design and Evidence schemas to runtime authority while
+  retaining alpha inputs as migration-only.
+- Bound natural-language UI/UX hook context with lazy contract documents and
+  require contract-bound route, state, and viewport PNG inventory.
+- Fail closed when renderer, auth, reviewer, or cleanup capabilities—or evidence
+  freshness and root provenance—cannot be proved, with schema/runtime parity and
+  adversarial real-surface coverage.
+
 ## 0.3.35 - 2026-07-26 — parity runtime release preparation
 
 - Restore and enroll the complete `litwork`, rules-engine, plan-scaffolding,

package/README.md

@@ -9,7 +9,7 @@
 </p>
 <p align="center">
   <img src="https://img.shields.io/badge/npm-litclaude--ai-cb3837" />
-  <img src="https://img.shields.io/badge/version-0.3.35-2ea44f" />
+  <img src="https://img.shields.io/badge/version-0.3.37-2ea44f" />
   <img src="https://img.shields.io/badge/Claude%20Code-plugin-blueviolet" />
   <img src="https://img.shields.io/badge/license-MIT-blue" />
 </p>
@@ -22,23 +22,16 @@
 > `litclaude@litclaude-ai`, so normal `claude` launches can load the
 > LitClaude skills and hooks without a long `--plugin-dir` command.
 
-This checkout is prepared as `litclaude-ai@0.3.35` for personal install
-convenience. The repo can remain quiet; preparing npm package metadata here does
-not imply public repo promotion, marketplace publication, or advertisement.
-The v0.3.35 release material restores the full `litwork`, rules-engine,
-plan-scaffolding, litgoal-drift, structural-search, comment-checker, and teammode
-surfaces; tracks their tests; and hash-pins shipped executable skill resources
-for installed tamper detection. Permission modes now mutate global Claude settings
-`permissions.allow` and `permissions.deny`: pre-existing entries are preserved, only
-entries LitClaude inserted are ownership-tracked, and only those inserted entries are
-removed on a later mode change or uninstall.
-Future package releases still require explicit user approval. The v0.3.34 release material adds packaged evidence-first UI/UX design intelligence, strict Design/Evidence/Review contracts, bounded PNG and TUI inspection, tier-correct independent review, and adversarial zero-false-PASS coverage. The v0.3.33 release material adds a quiet nonblocking npm update notice with strict cache, concurrency, timestamp, opt-out, and official-registry boundaries; preserves the installed permission mode and HUD accent on optionless updates; and makes source, install, installed-doctor, and Claude-details checks share the exact canonical skill catalog so missing packaged skills fail with `DOCTOR_FAIL`. The v0.3.32 release material adds the code-owned schema-3 start-work lifecycle with canonical bounded authority, explicit single-use resume grants, PreToolUse enforcement, root/subagent lane identity, progress-bound Stop continuation receipts, and fail-closed release boundaries. The v0.3.31 release material hardens litgoal creation, replacement, completion, and autoloop receipts while making visual QA browser ownership, bounded waits, blocked states, and cleanup evidence explicit. The v0.3.30 release material strengthens LitResearch with root-owned scientific evidence records, DOI/PDF lifecycle receipts, bounded route coverage, and a connection-pinned public-source reader with inert-content and secret-redaction guarantees. The v0.3.29 release material bundles exact-source `lit-handoff` and `lit-scientific-visualization` skills with Claude-native commands, exact-bare hook invocation, visible LITBURN banners, installed-payload integrity checks, and read-only scientific dependency diagnostics. The v0.3.28 release material requires explicit leading start-work invocations so diagnostic or copied mentions remain inert. The v0.3.27 release material adds a structured, animated five-stage installer with a success or failure receipt while preserving deterministic automation output. The v0.3.26 release material adds adaptive objective-achievable planning, draft-plan review, and orchestration readiness hardening. The v0.3.25 release material aligns portable details, public-read, litgoal status JSON, and evidence wording. The v0.3.24 release material aligns auxiliary skill inventories and advisory probes. The v0.3.23
-release materials inject the bundled skill bodies for bare `hyperplan`,
-`litresearch`, `lit research`, `init-deep`, and explicit `$start-work` prompt-hook routes while
-preserving the v0.3.19 adversarial planning skill, the v0.3.18 read-only `lit-recap` session recap surface, the v0.3.17 separate-worker native `/goal` launcher, the v0.3.16
-WSL2/Windows Terminal truecolor HUD gradient fix, the v0.3.15 Korean AI slop
-remover workflow, the v0.3.13 public-source reader runtime, litresearch
-activation/read-only polish, and installer permission-preference discipline.
+This checkout contains the current package source and local install metadata. That
+does not imply public promotion, marketplace publication, or advertisement. Package
+publication, version changes, tags, and releases always require separate explicit
+user approval. Current usage and safety behavior are documented here; cumulative
+release history belongs in [CHANGELOG.md](./CHANGELOG.md).
+
+Permission modes mutate global Claude settings `permissions.allow` and
+`permissions.deny`: pre-existing entries are preserved, only entries LitClaude
+inserted are ownership-tracked, and only those inserted entries are removed on a
+later mode change or uninstall.
 
 ## Features
 
@@ -48,22 +41,22 @@ activation/read-only polish, and installer permission-preference discipline.
   Claude Code's native command surface and do not double-activate the hook
 - **Deep interview mode** - type `/deep-interview` or `$deep-interview` to
   turn broad intent into a spec before planning or implementation
-- **v0.2.0 workflow parity** - adds `review-work` 5-lane review discipline and
+- **Workflow review discipline** - adds `review-work` 5-lane review discipline and
   litgoal runtime docs without claiming a package publication has completed
-- **v0.2.2 Dynamic workflow hardening** - adds stricter subagent assignment
+- **Dynamic workflow hardening** - adds stricter subagent assignment
   contracts, `start-work-next`, context-pressure resume guidance, precise
   mutated-file detection, and richer `workflow-check --json` readiness checks
-- **v0.3.10 build-decision gate** - a minimum-first check that runs before the
+- **Build-decision gate** - a minimum-first check that runs before the
   language gate and prefers the standard library, a native feature, or an
   installed dependency over hand-rolled code, with a reuse review after writing
-- **v0.3.11 native route hardening** - `lit goal` / `lit workflow` / `lit team`
+- **Native route hardening** - `lit goal` / `lit workflow` / `lit team`
   routes attempt native Claude surfaces honestly, report degraded mode when a
   host trigger is unavailable, and avoid phantom goal/workflow/team success
-- **v0.3.12 resilient public-source research** - `lit research`, `lit search`,
+- **Resilient public-source research** - `lit research`, `lit search`,
   and `lit query` use validator-first source checks, public API/feed preference,
   route traces, metadata fallback, and explicit stop reasons without crossing
   authentication, paywall, or private-data boundaries
-- **v0.3.13 public-source reader runtime** - `litclaude public-read <url> --json`
+- **Public-source reader runtime** - `litclaude public-read <url> --json`
   and MCP `public_source_read` provide a callable JS-only public page reader with
   SSRF/private-host guards, auth/paywall stop reasons, metadata/JSON-LD fallback,
   and route evidence for research lanes
@@ -83,13 +76,18 @@ activation/read-only polish, and installer permission-preference discipline.
 - **Claude skills** - a richer LitClaude-owned corpus: `programming`,
   `debugging`, `refactor`, `ai-slop-remover`, `remove-ai-slops`,
   `korean-ai-slop-remover`, `review-work`, `frontend-ui-ux`,
-  `git-master`, `comment-checker`, `rules`, `lsp`, `lsp-setup`, `litgoal`,
+  `git-master`, `comment-checker`, `lit-comprehend`, `rules`, `lsp`, `lsp-setup`, `litgoal`,
   `deep-interview`, `hyperplan`, `init-deep`, `litresearch`,
   `lit-plan`, `lit-recap`, `lit-loop`, `litwork`, `start-work`, `structural-search`,
   `teammode`, `lit-handoff`, `lit-scientific-visualization`, and `visual-qa`
+- **Canonical workflow families** - one Claude-native `autoresearch` skill routes
+  core/debug/fix/learn/plan/predict/reason/scenario/security/ship; one
+  `autoconference` skill routes core/analyze/debate/plan/resume/ship/survey and
+  fails with `BLOCKED_MULTI_AGENT_UNAVAILABLE` when root multi-agent capability is
+  absent; one task-local `wikify` skill routes init/ingest/query/save/lint
 - **Bundled owner-authorized skills** - `lit-handoff` reads the complete immutable
   four-file handoff source and supports the exact bare `handoff` prompt;
-  `lit-scientific-visualization` reads the complete immutable 16-file authored
+  `lit-scientific-visualization` reads its complete immutable authored
   source, resolves packaged scripts and assets, and reports missing Python or
   matplotlib as `DEGRADED` without silently installing dependencies
 - **Auxiliary Skill-discovery entries with chat routes** - `frontend-ui-ux`,
@@ -109,11 +107,31 @@ activation/read-only polish, and installer permission-preference discipline.
 - **Auxiliary workflow packs** - ships `programming/references`,
   `programming/scripts`, and `debugging/references` for deeper language and
   runtime guidance
-- **Interface reference packs** - ships `frontend-ui-ux/references` (fifteen
-  design decision documents, each routed from a table in its skill entrypoint)
-  and `visual-qa/references` (a per-channel capture playbook). Every reference
+- **Interface reference packs** - ships routed design decision documents under
+  `frontend-ui-ux/references` and `visual-qa/references` (a per-channel capture playbook). Every reference
   file is hash-pinned, and `npm run validate:plugin` fails when a shipped
-  reference is not reachable from its own entrypoint
+  reference is not reachable from its own entrypoint. Natural UI/UX activation
+  stays within 4096 UTF-8 bytes by injecting a synthetic bounded hook pointer, not the exact
+  `SKILL.md` body; each preserved dense contract is opened lazily from
+  `references/complete-contract.md`.
+- **Exact canonical frontend library** - the optional `design`, `designpowers`,
+  `perfection`, and `ui-ux-db` roots are byte-pinned by a path/size/SHA-256
+  manifest with exact legal companions and no-extra checks. They are inert data,
+  never executable authority, and remain independent of the unchanged
+  34-source/2,277-record normalized design-intelligence dataset. Legal companion
+  paths, source paths, sizes, hashes, and aggregate are independently pinned outside
+  the generated manifest; these clean legal files are scanned normally rather than
+  receiving the corpus exemption. The package guard compares size and SHA-256 for every
+  canonical frontend corpus file, legal companion, runtime closure, manifest, and family
+  adapter between its verified pre-pack snapshot and the produced temporary tarball; path
+  presence alone is insufficient. Each canonical verifier returns an immutable expected
+  file map bound to verified manifests or independent commitments, and capture checks its
+  bytes against that map so mutation in the verifier-to-capture interval fails. The same
+  map requires regular, non-executable canonical files in source, capture, and tar headers;
+  any execute bit fails, while secure non-executable modes such as `0600` remain valid.
+  Canonical frontend preflight applies an 8 MiB
+  per-file and 32 MiB aggregate read budget, and runtime closure preflight applies the
+  same 8 MiB per-file and 32 MiB aggregate budget before whole-file reads.
 - **Workflow agents** - planner, executor, verifier, reviewer, librarian, and QA runner
 - **Local marketplace registration** - Claude can resolve `claude plugin details litclaude@litclaude-ai`
 - **MCP and LSP helpers** - plugin-local stdio MCP plus TypeScript-family LSP doctor
@@ -149,13 +167,30 @@ package name as the registry package, older published builds such as
 `npx --yes litclaude-ai@0.1.11 install` could resolve the local same-name
 source checkout and fail with `sh: litclaude-ai: command not found`. This
 checkout intentionally does not track a `node_modules/.bin` shim. From a fresh
-directory, the normal install command works:
+directory, use `@latest` for the version currently published on npm:
+
+```bash
+cd /tmp
+npx --yes litclaude-ai@latest install
+```
+
+`0.3.37` is the version prepared by this checkout. Registry lookup is authoritative
+for whether that exact version can be installed:
+
+```bash
+npm view litclaude-ai@0.3.37 version
+```
+
+If that lookup returns 0.3.37, the exact install is available:
 
 ```bash
 cd /tmp
-npx --yes litclaude-ai@0.3.35 install
+npx --yes litclaude-ai@0.3.37 install
 ```
 
+Otherwise, wait for explicit human publication and repeat the lookup before using
+the exact version.
+
 Validate the installed plugin:
 
 ```bash
@@ -181,7 +216,7 @@ The installer also sets Claude Code's `statusLine` command to the packaged
 LitClaude HUD. A typical no-color render starts like:
 
 ```text
-[🔥LITCLAUDE v0.3.35] | O4.8 │ ctx [▎░░] 9%/1000k │ 5h [▏░] 4% ↻2h15m │ 1w [▊░] 35% ↻3d6h │ git main +3 ✓
+[🔥LITCLAUDE v0.3.37] | O4.8 │ ctx [▎░░] 9%/1000k │ 5h [▏░] 4% ↻2h15m │ 1w [▊░] 35% ↻3d6h │ git main +3 ✓
 ```
 
 The `↻` suffix is a compact rate-limit reset countdown. It is separated from
@@ -299,6 +334,12 @@ handoff
 /litclaude:lit-handoff
 lit-scientific-visualization
 /litclaude:lit-scientific-visualization <publication figure request>
+autoresearch <mode> <objective>
+/litclaude:autoresearch <mode> <objective>
+autoconference <mode> <objective>
+/litclaude:autoconference <mode> <objective>
+wikify <init|ingest|query|save|lint> <scope or question>
+/litclaude:wikify <init|ingest|query|save|lint> <scope or question>
 ```
 
 Expected behavior: LitClaude's prompt hook adds mode-specific `LITWORK MODE
@@ -358,7 +399,7 @@ synthesis. Optional display setup is `claude --teammate-mode auto` or
 `"teammateMode": "auto"`; if the env gate is absent, LitClaude says so and
 offers subagent/Dynamic workflow fallback without pretending teammates launched.
 
-For v0.2.2 Dynamic workflow hardening, `/dynamic-workflow` is the consolidated
+`/dynamic-workflow` is the consolidated
 bootstrap route for broad delegated work. It attempts native goal binding when a
 supported goal tool surface exists, reports degraded `BLOCKED:` mode when it
 does not, keeps `/goal` user-controlled, and maps subagent delegation to
@@ -420,7 +461,7 @@ instead of replaying stale blocking output. Subagent hooks
 bind root-session, child-lane, and Claude-owned worktree identity; active lanes
 gate completion.
 
-For v0.2.0 workflow parity, `review-work` is the dedicated review route. It is a
+`review-work` is the dedicated review route. It is a
 5-lane review: scope/diff verification, tests/evidence execution,
 package/payload and code quality, security/provenance, and real-surface/docs
 readiness. The lanes bind to
@@ -596,6 +637,21 @@ npm run qa:real-surface
 npm run pack:dry-run
 ```
 
+`npm run scan:legacy-tokens` enumerates the repository once, captures every
+enumerated file's exact bytes into an immutable in-memory snapshot, and scans
+only that captured snapshot. Its report includes the snapshot file count and a
+deterministic SHA-256 digest over the sorted inventory and bytes. A clean result
+is scoped to those captured bytes: it does not prove the mutable live tree
+remained clean after capture. Run the scanner again for a later tree state;
+package and tarball guards independently verify their exact payloads. The tarball
+guard compares every enrolled canonical file's size and SHA-256 with its verified
+pre-pack snapshot, including corpus, legal, runtime closure, manifest, and adapter
+files. Its immutable expected file map comes from the completed canonical verifiers,
+so the verifier-to-capture interval cannot silently redefine expected bytes. It also
+checks regular, non-executable source and tar semantics without requiring exact `0644`;
+secure `0600` remains valid. Capture is bounded to 8 MiB per file and 32 MiB in aggregate; either limit returns a typed
+failure instead of attempting an unbounded whole-repository allocation.
+
 `validate:plugin` and `qa:tmux` are local checks. If Claude Code is not
 available on the machine, the smoke harness records a controlled skip with the
 version probe evidence instead of failing mysteriously.

package/README_ko-KR.md

@@ -9,7 +9,7 @@
 </p>
 <p align="center">
   <img src="https://img.shields.io/badge/npm-litclaude--ai-cb3837" />
-  <img src="https://img.shields.io/badge/version-0.3.35-2ea44f" />
+  <img src="https://img.shields.io/badge/version-0.3.37-2ea44f" />
   <img src="https://img.shields.io/badge/Claude%20Code-plugin-blueviolet" />
   <img src="https://img.shields.io/badge/license-MIT-blue" />
 </p>
@@ -26,17 +26,14 @@
 > 설치되므로, 매번 긴 `--plugin-dir` 없이 일반 `claude` 실행에서
 > LitClaude skill과 hook을 불러올 수 있습니다.
 
-현재 checkout은 `litclaude-ai@0.3.35` 배포 준비용으로 정리되어 있습니다. 목적은
-다른 PC에서도 빠르게 설치하기 위한 개인용 package metadata를 갖추는 것입니다.
-npm package metadata를 준비했다고 해서 홍보, 공개 저장소 운영, Claude
-marketplace 등록을 의미하지는 않습니다. 새 버전 배포는 항상 별도의 명시적
-승인 후에 진행합니다. v0.3.34 release material은 packaged evidence-first UI/UX design intelligence, strict Design/Evidence/Review contract, bounded PNG/TUI inspection, tier-correct independent review, adversarial zero-false-PASS coverage를 추가합니다. v0.3.33 release material은 strict cache, concurrency, timestamp, opt-out, official-registry boundary를 갖춘 조용한 nonblocking npm update notice를 추가하고, optionless update에서 기존 permission mode와 HUD accent를 보존하며, source/install/installed doctor/Claude details가 정확한 canonical skill catalog를 공유해 packaged skill 누락 시 `DOCTOR_FAIL`로 실패하도록 합니다. v0.3.32 release material은 canonical bounded authority, single-use resume grant, PreToolUse enforcement, root/subagent lane identity, progress-bound Stop continuation receipt, fail-closed release boundary를 갖춘 code-owned schema-3 start-work lifecycle을 추가합니다. v0.3.31 release material은 litgoal 생성·교체·완료와 autoloop receipt를 강화하고 visual QA의 browser ownership, bounded wait, blocked state, cleanup evidence를 명확히 합니다. v0.3.30 release material은 root-owned scientific evidence record, DOI/PDF lifecycle receipt, bounded route coverage, connection-pinned public-source reader, inert-content 및 secret-redaction 보장을 포함하는 LitResearch 강화판입니다. v0.3.29 release material은 exact-source `lit-handoff`와 `lit-scientific-visualization` skill, Claude-native command, exact-bare hook invocation, visible LITBURN banner, installed-payload integrity check, read-only scientific dependency diagnostic을 함께 포함합니다. v0.3.28 release material은 start-work 호출을 prompt 선두의 명시적 invocation으로 제한해 진단문이나 복사된 언급을 비활성 상태로 유지합니다. v0.3.27 release material은 success/failure receipt와 deterministic automation output을 갖춘 5-stage animated installer를 추가합니다. v0.3.26 release material은 adaptive objective-achievable planning, draft-plan review, orchestration readiness hardening을 추가합니다. v0.3.25 release material은 portable details, public-read, litgoal status JSON, evidence wording을 정렬합니다. v0.3.24 release material은 auxiliary skill inventory와 advisory probe를 정렬합니다. v0.3.23 release material은 bare `hyperplan`,
-`litresearch`, `lit research`, `init-deep`, explicit `$start-work` prompt-hook route에
-bundled skill body를 주입하면서, v0.3.19 adversarial planning
-skill, v0.3.18 read-only `lit-recap` session recap surface, v0.3.17 별도 worker 기반 native `/goal`
-launcher와 v0.3.16 WSL2/Windows Terminal truecolor HUD gradient fix, v0.3.15
-Korean AI slop remover workflow, v0.3.13 public-source reader runtime,
-litresearch activation/read-only polish를 유지합니다.
+현재 checkout에는 현재 package source와 local install metadata가 있습니다. 이것은
+홍보, 공개 marketplace 등록, package publication을 의미하지 않습니다. publish, version
+변경, tag, release는 항상 별도의 명시적 사용자 승인이 필요합니다. 현재 사용법과 안전
+경계는 이 문서에서 설명하고, 누적 변경 이력은 [CHANGELOG.md](./CHANGELOG.md)에 둡니다.
+
+Permission mode는 global Claude settings의 `permissions.allow`와 `permissions.deny`를
+변경합니다. 기존 항목은 보존하고, LitClaude가 실제로 추가한 항목만 소유권을 기록하며,
+mode 변경이나 uninstall 때도 그 항목만 제거합니다.
 
 ## 기능
 
@@ -46,22 +43,22 @@ litresearch activation/read-only polish를 유지합니다.
   Claude Code native command surface에서 처리되어 hook이 중복 활성화하지 않습니다.
 - **Deep interview mode** - `/deep-interview` 또는 `$deep-interview`로 넓고
   모호한 의도를 먼저 spec으로 정리한 뒤 계획/구현으로 넘김
-- **v0.2.0 workflow parity** - package publish가 끝났다고 말하지 않으면서
+- **Workflow review discipline** - package publish가 끝났다고 말하지 않으면서
   `review-work` 5-lane review discipline과 litgoal runtime 문서를 추가
-- **v0.2.2 Dynamic workflow hardening** - 더 엄격한 subagent assignment
+- **Dynamic workflow hardening** - 더 엄격한 subagent assignment
   contract, `start-work-next`, context-pressure resume guidance, 정확한
   mutated-file 감지, 강화된 `workflow-check --json` readiness check를 추가
-- **v0.3.10 build-decision gate** - language gate 앞에서 minimum-first check를
+- **Build-decision gate** - language gate 앞에서 minimum-first check를
   돌려 표준 라이브러리·네이티브 기능·이미 설치된 의존성을 직접 작성보다 먼저
   쓰게 하고, 작성 후 reuse review를 추가
-- **v0.3.11 native route hardening** - `lit goal` / `lit workflow` / `lit team`
+- **Native route hardening** - `lit goal` / `lit workflow` / `lit team`
   route가 Claude native surface를 정직하게 시도하고, host trigger가 없으면
   degraded mode를 보고하며 가짜 goal/workflow/team 성공을 피함
-- **v0.3.12 public-source research 강화** - `lit research`, `lit search`,
+- **Public-source research 강화** - `lit research`, `lit search`,
   `lit query`가 public API/feed 우선, validator-first 검증, route trace,
   metadata fallback, 명확한 stop reason을 사용하고 인증/페이월/개인 데이터
   경계를 넘지 않음
-- **v0.3.13 public-source reader runtime** - `litclaude public-read <url> --json`와
+- **Public-source reader runtime** - `litclaude public-read <url> --json`와
   MCP `public_source_read`가 SSRF/private-host guard, auth/paywall stop reason,
   metadata/JSON-LD fallback, route evidence를 제공하는 JS-only public page reader를 추가
 - **5-lane review** - `/review-work`가 scope/diff verification,
@@ -80,12 +77,17 @@ litresearch activation/read-only polish를 유지합니다.
 - **Claude skills** - LitClaude 스타일을 더 풍부하게 옮긴 corpus:
   `programming`, `debugging`, `refactor`, `ai-slop-remover`, `remove-ai-slops`,
   `korean-ai-slop-remover`, `review-work`, `frontend-ui-ux`,
-  `git-master`, `comment-checker`, `rules`, `lsp`, `lsp-setup`, `litgoal`,
+  `git-master`, `comment-checker`, `lit-comprehend`, `rules`, `lsp`, `lsp-setup`, `litgoal`,
   `deep-interview`, `hyperplan`, `init-deep`, `litresearch`,
   `lit-plan`, `lit-recap`, `lit-loop`, `start-work`, `lit-handoff`,
-  `lit-scientific-visualization`, `visual-qa`
+  `structural-search`, `teammode`, `lit-scientific-visualization`, `visual-qa`
+- **Canonical workflow family** - Claude-native `autoresearch`는
+  core/debug/fix/learn/plan/predict/reason/scenario/security/ship을,
+  `autoconference`는 core/analyze/debate/plan/resume/ship/survey를 라우팅하며 root
+  multi-agent capability가 없으면 `BLOCKED_MULTI_AGENT_UNAVAILABLE`로 멈춥니다.
+  task-local `wikify`는 init/ingest/query/save/lint를 라우팅합니다.
 - **원본 보존 bundled skills** - `lit-handoff`는 승인된 4개 handoff 원본
-  파일을, `lit-scientific-visualization`은 승인된 authored 원본 16개를
+  파일을, `lit-scientific-visualization`은 승인된 authored 원본을
   byte-identical하게 포함합니다. Python 또는 matplotlib이 없으면 자동
   설치하지 않고 해당 capability를 `DEGRADED`로 보고합니다.
 - **Auxiliary Skill-discovery entries + chat route** - `frontend-ui-ux`,
@@ -107,11 +109,41 @@ litresearch activation/read-only polish를 유지합니다.
 - **Auxiliary workflow packs** - `programming/references`,
   `programming/scripts`, `debugging/references`까지 포함해서 언어별,
   런타임별 세부 지침을 보존
-- **Interface reference packs** - `frontend-ui-ux/references`(디자인 결정
-  문서 15개, 각 skill entrypoint의 라우터 표에서 바로 찾아갈 수 있음)와
+- **Interface reference packs** - 라우팅된 디자인 결정 문서가 있는
+  `frontend-ui-ux/references`와
   `visual-qa/references`(채널별 capture playbook)를 포함합니다. 모든 reference
   파일은 hash-pinned이고, entrypoint에서 도달할 수 없는 reference가 있으면
-  `npm run validate:plugin`이 실패합니다.
+  `npm run validate:plugin`이 실패합니다. canonical manifest가 shipped resource 무결성을 검증합니다. 자연어 UI/UX hook은 정확한 `SKILL.md` 본문이 아니라 synthetic bounded hook pointer를
+  주입하고 상세 `references/complete-contract.md`는 필요할 때만 읽습니다. prompt-time rule을 합친 최종
+  `additionalContext`도 UTF-8 기준 **4096 바이트 이하**입니다.
+- **Exact canonical frontend library** - `design`, `designpowers`, `perfection`,
+  `ui-ux-db`는 path/size/SHA-256 manifest, exact legal companion, no-extra gate로
+  byte-pinned됩니다. 모두 inert data이며 실행 권한이 아닙니다. 기존
+  34-source/2,277-record normalized design-intelligence dataset은 변경 없이 독립적으로 유지됩니다.
+  Legal 파일의 path, source path, size, hash, aggregate는 generated manifest 외부에서
+  독립적으로 고정되며, 이 파일들은 corpus 예외를 받지 않고 일반 검사됩니다.
+  Package guard는 검증된 pre-pack snapshot과 생성된 임시 tarball을 비교해 canonical frontend
+  corpus, legal companion, runtime closure, manifest, family adapter 전 파일의 size와 SHA-256을
+  확인합니다. Path 존재만으로는 통과할 수 없습니다. 각 verifier는 검증된 manifest 또는
+  독립 commitment에 묶인 immutable expected file map을 반환하며, capture는 바이트를 이 map과
+  대조하므로 verifier-to-capture interval의 변경도 실패합니다. 같은 map은 source, capture,
+  tar header의 regular non-executable semantics를 요구합니다. Execute bit가 하나라도 있으면
+  실패하지만 exact `0644`는 요구하지 않으므로 secure `0600`은 유효합니다.
+  Canonical frontend preflight는 whole-file read 전에 파일당 8 MiB와 전체 32 MiB를,
+  runtime closure preflight도 파일당 8 MiB와 전체 32 MiB를 한도로 적용합니다.
+- **Beta Design/Evidence authority** - UI 작성의 권위 스키마는
+  `litfamily.design-contract/v1beta1`, 시각 증거의 권위 스키마는
+  `litfamily.evidence-manifest/v1beta1`입니다. alpha는 migration-only이며
+  `LEGACY_SCHEMA_V1ALPHA1`/`evidence_eligible: false`로 남습니다. Beta smoke evidence는
+  beta contract에 선언된 primary route, critical interaction, 적용 가능한 negative/empty state,
+  최소/최대 viewport에만
+  연결되고, viewport마다 contract의 width/height와 일치하는 PNG가 필요합니다. 호환되지
+  않는 viewport를 PNG 하나로 충족하거나 선언되지 않은 inventory를 넣으면 `PASS`가
+  아니라 `EVIDENCE_CONTRACT_INVENTORY_MISMATCH` 또는
+  `EVIDENCE_VIEWPORT_BINDING_INVALID`입니다. renderer/auth/reviewer/cleanup capability가
+  없으면 정확한 `BLOCKED_*`를 유지하며 `PASS`로 낮추지 않습니다. 공개 JSON, 갱신한 mtime,
+  호출자가 넣은 current hash/revision은 host capture provenance를 증명하지 못하므로 현재 pathname
+  CLI는 `BLOCKED_EVIDENCE_FRESHNESS_UNPROVEN`과 `BLOCKED_EVIDENCE_ROOT_UNPROVEN`을 반환합니다.
 - **워크플로우 agents** - planner, executor, verifier, reviewer, librarian, QA runner
 - **Local marketplace 등록** - `claude plugin details litclaude@litclaude-ai`에서 inventory 확인
 - **MCP와 LSP helper** - plugin-local stdio MCP와 TypeScript 계열 LSP doctor
@@ -144,13 +176,30 @@ plain `INSTALL_STEP` line을 유지하며 `--dry-run`은 계속 side-effect-free
 `npx --yes litclaude-ai@0.1.11 install` 같은 이전 published build가 local
 checkout을 먼저 해석해서 `sh: litclaude-ai: command not found`로 실패할 수
 있었습니다. 이 checkout은 `node_modules/.bin` shim을 의도적으로 추적하지
-않습니다. 새 폴더에서는 일반 설치 명령이 정상 동작합니다.
+않습니다. 새 폴더에서는 npm에 현재 공개된 버전을 가리키는 `@latest`를 사용합니다.
+
+```bash
+cd /tmp
+npx --yes litclaude-ai@latest install
+```
+
+`0.3.37`은 이 checkout에서 준비 중인 version입니다. exact version의 설치 가능
+여부에는 registry 조회 결과를 authoritative한 기준으로 사용합니다.
+
+```bash
+npm view litclaude-ai@0.3.37 version
+```
+
+조회 결과가 0.3.37이면 exact install을 사용할 수 있습니다.
 
 ```bash
 cd /tmp
-npx --yes litclaude-ai@0.3.35 install
+npx --yes litclaude-ai@0.3.37 install
 ```
 
+그렇지 않으면 명시적인 human publication을 기다립니다. 게시 후 조회를 다시 실행해
+candidate가 반환되는지 확인한 뒤 exact version을 사용합니다.
+
 설치 상태를 확인합니다.
 
 ```bash
@@ -176,7 +225,7 @@ installer는 Claude Code의 `statusLine` command도 packaged LitClaude HUD로
 설정합니다. 색상을 제거한 예시는 다음처럼 시작합니다.
 
 ```text
-[🔥LITCLAUDE v0.3.35] | O4.8 │ ctx [▎░░] 9%/1000k │ 5h [▏░] 4% ↻2h15m │ 1w [▊░] 35% ↻3d6h │ git main +3 ✓
+[🔥LITCLAUDE v0.3.37] | O4.8 │ ctx [▎░░] 9%/1000k │ 5h [▏░] 4% ↻2h15m │ 1w [▊░] 35% ↻3d6h │ git main +3 ✓
 ```
 
 `↻` 표시는 rate-limit reset까지 남은 시간을 짧게 보여주는 countdown입니다.
@@ -287,6 +336,12 @@ handoff
 /litclaude:lit-handoff
 lit-scientific-visualization
 /litclaude:lit-scientific-visualization <출판용 figure 요청>
+autoresearch <mode> <objective>
+/litclaude:autoresearch <mode> <objective>
+autoconference <mode> <objective>
+/litclaude:autoconference <mode> <objective>
+wikify <init|ingest|query|save|lint> <scope 또는 question>
+/litclaude:wikify <init|ingest|query|save|lint> <scope 또는 question>
 ```
 
 기대 동작은 LitClaude prompt hook이 자연어 LIT route별 mode-specific
@@ -330,7 +385,7 @@ capability boundary와 ledger 기반 ready-to-paste condition은 clean-room
 worktree를 필요로 하면 `EnterWorktree`를 사용합니다. CLI surface만 가능한 새 격리
 lane에서는 `claude --worktree <short-name> --tmux`를 사용합니다.
 
-v0.2.2 Dynamic workflow hardening에서 `/dynamic-workflow`는 넓은 위임 작업을
+`/dynamic-workflow`는 넓은 위임 작업을
 위한 통합 bootstrap route입니다. `/goal`은 user-controlled 상태로 두고,
 model-facing goal tools가 노출되면 확인하며, 필요하면 정확한 fallback
 `/goal <completion condition>`만 제안합니다. subagent delegation은
@@ -355,7 +410,7 @@ litclaude-ai start-work-next --session-id <claude-session> --json
 plan, ledger path, 첫 번째 unchecked top-level task를 출력합니다. Plan이 끝났으면
 `{"status":"idle","directive":null}`을 반환합니다.
 
-v0.2.0 workflow parity에서 `review-work`는 전용 review route입니다. 5-lane
+`review-work`는 전용 review route입니다. 5-lane
 review는 scope/diff verification, tests/evidence execution,
 package/payload and code quality, security/provenance,
 real-surface/docs readiness를 포함합니다. 각 lane은
@@ -529,6 +584,20 @@ npm run qa:real-surface
 npm run pack:dry-run
 ```
 
+`npm run scan:legacy-tokens`는 저장소를 한 번 열거하고, 열거된 모든 파일의
+정확한 바이트를 변경 불가능한 인메모리 스냅샷으로 캡처한 뒤 그 스냅샷만
+검사합니다. 보고서에는 스냅샷 파일 수와 정렬된 인벤토리 및 바이트의
+결정론적 SHA-256 다이제스트가 포함됩니다. clean 결과는 캡처된 바이트에만
+적용되며, 캡처 이후 mutable 라이브 트리가 계속 clean이었다는 사실을
+입증하지 않습니다. 이후 트리 상태는 스캐너를 다시 실행해야 하며 package와
+tarball guard는 각자의 정확한 payload를 독립적으로 검증합니다. Corpus, legal, runtime
+closure, manifest, adapter로 등록된 모든 canonical 파일의 size와 SHA-256을 검증된
+pre-pack snapshot과 비교합니다. Immutable expected file map은 완료된 verifier 결과에서
+오므로 verifier-to-capture interval에서 기대 바이트를 바꿀 수 없습니다. Source와 tar의
+regular non-executable semantics도 검사하며 exact `0644` 대신 secure `0600`을 허용합니다. 캡처 한도는
+파일당 8 MiB, 전체 32 MiB이며, 한도를 넘으면 저장소 전체를 무제한으로
+할당하지 않고 typed failure를 반환합니다.
+
 `validate:plugin`과 `qa:tmux`는 로컬 검증용입니다. Claude Code가 없는
 환경에서는 smoke harness가 모호한 실패 대신 controlled skip과 version
 probe evidence를 남깁니다.

package/RELEASE_CHECKLIST.md

@@ -1,6 +1,13 @@
 # LitClaude Release Checklist
 
-Status: `litclaude-ai@0.3.35` is the current release candidate — restored
+Status: `litclaude-ai@0.3.37` is the current release candidate — exact canonical
+frontend corpus plus Claude-native `autoresearch`, `autoconference`, and `wikify`
+workflow-family integration. It byte-pins the frontend library, legal companions,
+family source closures, and adapters through independent commitments and package
+verification. It also carries G8 UI/UX capability and evidence hardening with
+beta Design/Evidence schema authority, bounded lazy hook context, contract-bound
+PNG inventory, and fail-closed
+provenance, freshness, root, and capability blockers. It also carries restored
 litwork, rules-engine, plan-scaffolding, litgoal-drift, structural-search,
 comment-checker, and teammode surfaces with tracked regression tests and
 installed tamper coverage, plus packaged evidence-first UI/UX design intelligence, strict Design/Evidence/Review
@@ -38,9 +45,9 @@ side-effect-free, the launcher starts only a separate Claude Code
 print/background worker, and the release preserves the Korean polishing
 command, strict multi-agent review pipeline, fidelity guardrails, package
 hygiene checks, native route gates, and safe start-work handoff behavior.
-`package.json` is aligned to `0.3.35`,
-`plugins/litclaude/.claude-plugin/plugin.json` is aligned to `0.3.35`, and the
-plugin-local MCP server reports `0.3.35`.
+`package.json` is aligned to `0.3.37`,
+`plugins/litclaude/.claude-plugin/plugin.json` is aligned to `0.3.37`, and the
+plugin-local MCP server reports `0.3.37`.
 
 This release carries the v0.2.2 Dynamic workflow hardening surfaces:
 `/dynamic-workflow`, `workflow-check --json`, native `/goal` fallback guidance,
@@ -63,7 +70,7 @@ selection, `0.1.16` added rate-limit reset countdown and deep-interview install
 guidance, and `0.1.18` covered compact HUD bars and broader brand accent
 behavior.
 
-## Verified Locally
+## Required Local Verification
 
 - `npm test`
 - `npm run validate:plugin`
@@ -92,6 +99,27 @@ Use this track when testing from the current checkout:
 
 No npm publication is required for this track.
 
+## v0.3.37 Canonical Frontend Corpus and Workflow-Family Gates
+
+Before requesting publication approval, confirm the `design`, `designpowers`,
+`perfection`, and `ui-ux-db` frontend roots match their path/size/SHA-256
+manifest, exact legal companions, independent commitments, no-extra inventory,
+regular non-executable modes, and bounded read budgets. The unchanged
+34-source/2,277-record normalized design dataset remains independently verified.
+
+Confirm the complete `autoresearch`, `autoconference`, and `wikify` source
+closures match their canonical manifest and independent tree commitments. Their
+Claude-native skills and commands must expose the documented mode inventories,
+safe leading bare and dollar hook routes, inert near misses, explicit publication
+boundaries, and `BLOCKED_MULTI_AGENT_UNAVAILABLE` when an autoconference cannot
+use root multi-agent capability.
+
+Run `npm run check:runtime-closures`, `npm run check:skill-resources`,
+`npm run validate:plugin`, and `npm run pack:payload-guard`. The payload guard
+must compare the verified source snapshot with the temporary tarball by size,
+SHA-256, and executable semantics for every canonical corpus file, legal
+companion, source-closure file, manifest, and family adapter.
+
 ## v0.3.34 Evidence-First UI/UX and Visual QA Gates
 
 Before requesting publication approval, confirm `frontend-ui-ux` ships the
@@ -127,7 +155,7 @@ validator on purpose, so a manifest remains checkable from a partial install and
 two independent readings of one shape can disagree visibly.
 
 Confirm the blocker vocabulary is complete and honest. `BLOCKED_AUTH_UNAVAILABLE`,
-`BLOCKED_CAPTURE_UNAVAILABLE`, `BLOCKED_INDEPENDENT_REVIEW_UNAVAILABLE`,
+`BLOCKED_RENDERER_UNAVAILABLE`, `BLOCKED_INDEPENDENT_REVIEW_UNAVAILABLE`,
 `BLOCKED_TEST_ACCOUNT_UNSAFE`, `BLOCKED_RENDERER_OWNERSHIP_UNVERIFIED`,
 `BLOCKED_EVIDENCE_STALE`, `BLOCKED_EVIDENCE_FUTURE`, and
 `BLOCKED_CLEANUP_INCOMPLETE` must each be emittable by the shipped validator,
@@ -252,9 +280,9 @@ checkout and from an isolated install of the packed tarball:
 Before requesting publication approval, confirm these artifacts from the current
 checkout:
 
-- `package.json` version is `0.3.35`.
-- `plugins/litclaude/.claude-plugin/plugin.json` version is `0.3.35`.
-- `plugins/litclaude/bin/litclaude-mcp.js` reports server version `0.3.35`.
+- `package.json` version is `0.3.37`.
+- `plugins/litclaude/.claude-plugin/plugin.json` version is `0.3.37`.
+- `plugins/litclaude/bin/litclaude-mcp.js` reports server version `0.3.37`.
 - Prompt-hook tests cover bundled `SKILL.md` body injection for bare `hyperplan`, `litresearch`, `lit research`, `init-deep`, and explicit leading `$start-work`; diagnostic/copy mentions stay inert while leading natural-language `lit start work` stays BLOCKED.
 - `lit search` and `lit query` route to `/litclaude:litresearch` without activating on slash mentions, code spans, or non-lit prompts.
 - Litresearch web lanes require public API/feed preference, validator-first checks, route traces, prompt-injection quarantine, and honest auth/paywall/private-data stop reasons.