ltcai 4.6.1 → 4.7.2

package/README.md

@@ -7,56 +7,70 @@
 [![CI Status](https://github.com/TaeSooPark-PTS/LatticeAI/actions/workflows/ci.yml/badge.svg?branch=main)](https://github.com/TaeSooPark-PTS/LatticeAI/actions/workflows/ci.yml)
 [![License](https://img.shields.io/badge/license-MIT-green)](LICENSE)
 
-**Lattice AI v4.6.1 is the local-first Living Brain release refresh.** It keeps
-the v4 Brain Core, StorageEngine, FastAPI localhost API, Tauri shell, backup,
-restore, model runtime, graph, and portability architecture intact while making
-the release/publishable version `4.6.1` after the v4.6.0 PyPI immutability
-block.
-
-The product opens with a private local profile, studies the computer, recommends
-models, asks before install/download/load work, then lands in Brain Chat. The
-graph is still real and searchable, but it now grows out of the Brain as the
-deepest exploration layer instead of acting like a separate dashboard.
+**Lattice AI v4.7.2 is the local-first Living Brain intuitive UX release.**
+It keeps the v4 Brain Core, StorageEngine, FastAPI localhost API, Tauri shell,
+backup, restore, model runtime, graph, and portability architecture intact while
+making the normal user surface easier for non-technical users: safer login,
+clearer model setup, one-click Memory/Topic/Relationship/Graph views, and
+visible feedback when conversation becomes Brain memory. Admin logs, role
+permissions, security events, retention posture, and Brain operations remain in
+the separate Admin Console.
+
+The product opens by stating the core reason to exist: models will change, but
+the user's knowledge should not. It creates a private local profile, studies the
+computer, recommends a replaceable model voice, asks before install/download/load
+work, then lands in Brain Chat with starter prompts for durable decisions,
+projects, documents, and context. The Brain home now gives direct buttons for
+memory, topics, relationships, and the full graph, so the graph is still real
+and searchable without becoming the first screen or a separate dashboard.
 
 External package registries are owner-published and can lag behind this GitHub
-Release. Release uploads must use the exact v4.6.1 artifact filenames below.
+Release. Release uploads must use the exact v4.7.2 artifact filenames below.
 
 ## Living Brain Flow
 
 ### 1. Login
 
 First launch opens to Login only. The local profile is the beginning of the
-Brain, not a dashboard, graph, or setup grid.
+Brain, not a dashboard, graph, or setup grid. The first screen frames Lattice as
+a durable knowledge home where models are replaceable and ownership stays with
+the user. v4.7.2 also prevents an email typo or wrong saved-user password from
+silently creating a new empty Brain.
 
-![Login](output/release/v4.6.1/screenshots/01-login.png)
+![Login](output/release/v4.7.2/screenshots/01-login.png)
 
 ### 2. Environment Analysis
 
 Lattice reads the machine locally and summarizes what kind of Brain this
 computer can support.
 
-![Environment Analysis](output/release/v4.6.1/screenshots/02-environment-analysis.png)
+![Environment Analysis](output/release/v4.7.2/screenshots/02-environment-analysis.png)
 
 ### 3. Recommended Models
 
 The model step is a short recommendation list. It avoids catalog noise and keeps
-runtime/install details behind clear user consent.
+runtime/install details behind clear user consent. Users who do not know which
+model to choose can start with the recommended model in one click.
 
-![Recommended Models](output/release/v4.6.1/screenshots/03-recommended-models.png)
+![Recommended Models](output/release/v4.7.2/screenshots/03-recommended-models.png)
 
 ### 4. Install And Load
 
 The install screen keeps consent visible and shows install, download, validate,
-and load progress. No model download or runtime install starts silently.
+and load progress. No model download or runtime install starts silently, and the
+screen explains that large downloads may take minutes without inventing fake ETA
+data.
 
-![Install and Load](output/release/v4.6.1/screenshots/04-install-load-progress.png)
+![Install and Load](output/release/v4.7.2/screenshots/04-install-load-progress.png)
 
 ### 5. Brain Chat
 
 After setup, the home experience is the living Brain plus conversation. The Brain
-stays present while the user types, recalls context, and receives responses.
+stays present while the user types, recalls context, and receives responses. The
+home now includes a compact Brain overview for recent memories, older memories,
+and major topics, plus saved-to-memory feedback after chat.
 
-![Brain Chat Home](output/release/v4.6.1/screenshots/05-brain-chat-home.png)
+![Brain Chat Home](output/release/v4.7.2/screenshots/05-brain-chat-home.png)
 
 ## Brain Depths
 
@@ -65,22 +79,26 @@ while revealing more structure.
 
 | Depth | Experience | Evidence |
 | --- | --- | --- |
-| Level 1 | Living Brain presence | ![Living Brain Level 1](output/release/v4.6.1/screenshots/06-living-brain-level-1.png) |
-| Level 2 | Memory Layer | ![Memory Layer](output/release/v4.6.1/screenshots/07-memory-layer.png) |
-| Level 3 | Knowledge Layer | ![Knowledge Layer](output/release/v4.6.1/screenshots/08-knowledge-layer.png) |
-| Level 4 | Relationship Layer | ![Relationship Layer](output/release/v4.6.1/screenshots/09-relationship-layer.png) |
-| Level 5 | Knowledge Graph with nodes, edges, search, and focus detail | ![Knowledge Graph Layer](output/release/v4.6.1/screenshots/10-knowledge-graph-layer.png) |
+| Level 1 | Living Brain presence | ![Living Brain Level 1](output/release/v4.7.2/screenshots/06-living-brain-level-1.png) |
+| Level 2 | Memory Layer | ![Memory Layer](output/release/v4.7.2/screenshots/07-memory-layer.png) |
+| Level 3 | Knowledge Layer | ![Knowledge Layer](output/release/v4.7.2/screenshots/08-knowledge-layer.png) |
+| Level 4 | Relationship Layer | ![Relationship Layer](output/release/v4.7.2/screenshots/09-relationship-layer.png) |
+| Level 5 | Knowledge Graph with nodes, edges, search, and focus detail | ![Knowledge Graph Layer](output/release/v4.7.2/screenshots/10-knowledge-graph-layer.png) |
 
 Walkthrough:
 
-![v4.6.1 Living Brain walkthrough](output/release/v4.6.1/gifs/v4.6.1-living-brain-walkthrough.gif)
+![v4.7.2 Living Brain walkthrough](output/release/v4.7.2/gifs/v4.7.2-living-brain-walkthrough.gif)
 
 Model setup status evidence:
 
-![Model setup status](output/release/v4.6.1/screenshots/11-model-setup-status.png)
+![Model setup status](output/release/v4.7.2/screenshots/11-model-setup-status.png)
+
+Separate admin console evidence:
+
+![Admin Console](output/release/v4.7.2/screenshots/12-admin-console.png)
 
 Screenshot index and capture notes:
-[output/release/v4.6.1/SCREENSHOT_INDEX.md](output/release/v4.6.1/SCREENSHOT_INDEX.md)
+[output/release/v4.7.2/SCREENSHOT_INDEX.md](output/release/v4.7.2/SCREENSHOT_INDEX.md)
 
 ## Architecture At A Glance
 
@@ -96,12 +114,21 @@ Screenshot index and capture notes:
 - **Storage**: `StorageEngine` abstraction with SQLite default and optional
   PostgreSQL/pgvector scale mode.
 - **Portability**: encrypted `.latticebrain` archives plus backup, restore,
-  inspect, verify, import dry-run, and confirmed restore/import flows.
+  inspect, verify, import dry-run, and confirmed restore/import flows. The
+  Brain home keeps conversation first and surfaces the everyday "Care for my
+  Brain" ownership path as a collapsed control with export, backup, archive,
+  inspect, and restore preview actions while keeping destructive confirmed
+  restore in Settings. Restore operations create pre-restore backups and roll
+  back failed DB/blob swaps so the current Brain is not left half-restored.
 - **Privacy**: local-first and private-first by default. Cloud models, Telegram,
   Brain Network, Docker/Postgres setup, model downloads, and update checks are
   opt-in paths.
+- **Admin separation**: user chat and Brain ownership stay in the main Brain
+  surface; user directory, audit logs, security events, policies, and index
+  rebuild controls live under the separate `#/admin` console. Admin history,
+  audit, stats, and sensitivity reads honor the active workspace when present.
 
-See [ARCHITECTURE.md](ARCHITECTURE.md) for the detailed v4.6.1 architecture.
+See [ARCHITECTURE.md](ARCHITECTURE.md) for the detailed v4.7.2 architecture.
 
 ## Installation
 
@@ -133,13 +160,13 @@ pip install "ltcai[local]"
 
 ## Release Artifacts
 
-Validated v4.6.1 artifacts:
+Validated v4.7.2 artifacts:
 
-- `dist/ltcai-4.6.1-py3-none-any.whl`
-- `dist/ltcai-4.6.1.tar.gz`
-- `ltcai-4.6.1.tgz`
-- `dist/ltcai-4.6.1.vsix`
-- `src-tauri/target/release/bundle/dmg/Lattice AI_4.6.1_aarch64.dmg`
+- `dist/ltcai-4.7.2-py3-none-any.whl`
+- `dist/ltcai-4.7.2.tar.gz`
+- `ltcai-4.7.2.tgz`
+- `dist/ltcai-4.7.2.vsix`
+- `src-tauri/target/release/bundle/dmg/Lattice AI_4.7.2_aarch64.dmg`
 
 Attach only those exact files to the GitHub Release. Do not upload `dist/*`.
 
@@ -168,7 +195,7 @@ npm run test:unit
 npm run test:integration
 npm run test:visual
 npm run desktop:tauri:check
-node scripts/run_python.mjs scripts/wheel_smoke.py --wheel dist/ltcai-4.6.1-py3-none-any.whl
+node scripts/run_python.mjs scripts/wheel_smoke.py --wheel dist/ltcai-4.7.2-py3-none-any.whl
 npm pack --dry-run
 npm run docs:check-links
 ```
@@ -184,13 +211,15 @@ npm run docs:check-links
 - Agent/workflow simulation without a loaded LLM is deterministic and does not call a model.
   It is labeled as LLM-free/model-free rather than presented as autonomous model
   success.
-- Historical artifacts can remain in `dist/`; uploads must use exact v4.6.1
+- Historical artifacts can remain in `dist/`; uploads must use exact v4.7.2
   filenames.
 
 ## Release History
 
 | Version | Theme |
 | --- | --- |
+| 4.7.2 | Intuitive Brain UX Release: adds role permissions, audit search/severity filters, log retention status, and separated Admin Console data loading while keeping the user Brain simple |
+| 4.7.0 | Admin Separation Release: added the separate Admin Console for users/logs/security/Brain operations, refreshed screenshots/GIFs, synchronized release docs, and built exact v4.7.0 artifacts |
 | 4.6.1 | Living Brain Release Refresh: publishable version bump after v4.6.0 PyPI immutability, refreshed README/screenshots/GIFs, synchronized release docs, and exact v4.6.1 artifacts |
 | 4.6.0 | Living Brain Experience: made Brain plus conversation the home product, added an animated living Brain presence, and moved graph exploration to the deepest intentional layer |
 | 4.5.1 | Product Reimagining RC: replaced the desktop shell, navigation model, onboarding journey, first-viewport hierarchy, and visual system while preserving capabilities and local-first architecture |
@@ -208,15 +237,20 @@ npm run docs:check-links
 
 ## Current Documentation
 
-- [ARCHITECTURE.md](ARCHITECTURE.md) - v4.6.1 architecture.
+- [ARCHITECTURE.md](ARCHITECTURE.md) - v4.7.2 architecture.
+- [docs/PRODUCT_DIRECTION_REVIEW.md](docs/PRODUCT_DIRECTION_REVIEW.md) -
+  Brain-first product direction review.
 - [FEATURE_STATUS.md](FEATURE_STATUS.md) - current feature status and historical
   status ledger.
 - [RELEASE_NOTES.md](RELEASE_NOTES.md) - release notes index.
-- [RELEASE_NOTES_v4.6.1.md](RELEASE_NOTES_v4.6.1.md) - v4.6.1 release refresh notes.
+- [RELEASE_NOTES_v4.7.2.md](RELEASE_NOTES_v4.7.2.md) - v4.7.2 intuitive Brain UX release notes.
+- [RELEASE_NOTES_v4.6.1.md](RELEASE_NOTES_v4.6.1.md) - v4.6.1 release refresh history.
 - [RELEASE_NOTES_v4.6.0.md](RELEASE_NOTES_v4.6.0.md) - v4.6.0 Living Brain history.
 - [RELEASE.md](RELEASE.md) - release checklist and exact artifact guidance.
 - [SECURITY.md](SECURITY.md) - security posture.
 - [docs/CHANGELOG.md](docs/CHANGELOG.md) - changelog.
+- [docs/V4_7_1_ADMIN_OPERATIONS_REPORT.md](docs/V4_7_1_ADMIN_OPERATIONS_REPORT.md) - v4.7.2 admin operations report.
+- [docs/V4_7_0_ADMIN_SEPARATION_REPORT.md](docs/V4_7_0_ADMIN_SEPARATION_REPORT.md) - v4.7.0 admin separation history.
 - [docs/V4_6_1_RELEASE_REFRESH_REPORT.md](docs/V4_6_1_RELEASE_REFRESH_REPORT.md) - v4.6.1 release refresh report.
 - [docs/V4_6_0_LIVING_BRAIN_EXPERIENCE_REPORT.md](docs/V4_6_0_LIVING_BRAIN_EXPERIENCE_REPORT.md) - v4.6.0 Living Brain design notes.
 

package/docs/CHANGELOG.md

@@ -3,6 +3,131 @@
 The top entry is the current release-preparation target. Older entries are
 historical and may describe behavior as it existed at that release.
 
+## [4.7.2] - 2026-06-14
+
+> Intuitive Brain UX Release. v4.7.2 makes the Living Brain easier for
+> non-technical users: first-run login is safer, recommended model setup is
+> one-click, Brain memory/topic/relationship/graph views are directly visible,
+> and conversation shows saved-to-memory feedback while Admin remains separate.
+
+### Changed
+
+- Prevented saved-user email mismatch or wrong saved-user password from silently
+  creating a new empty Brain during first-run login.
+- Added a primary `추천대로 시작하기` model recommendation path and clearer large
+  model download messaging without fake ETA.
+- Added visible `기억 보기`, `주제 보기`, `관계 보기`, and `그래프로 보기` actions
+  on the Brain surface so users can open the desired Brain depth directly.
+- Added a Brain overview panel with recent memories, older memories, major
+  topics, and saved-to-memory feedback after conversation.
+- Updated first-run and empty-Brain copy toward plain user language while
+  keeping the graph as the deepest advanced layer.
+- Bumped synchronized package/runtime/static versions to `4.7.2`, including
+  Python metadata, npm package metadata, VSIX metadata, Tauri metadata,
+  `latticeai`, `lattice_brain`, runtime constants, and static asset metadata.
+- Refreshed README, release notes, architecture, feature status, security,
+  recovery notes, VS Code extension docs, release report, and release evidence
+  paths for v4.7.2.
+
+### Preserved
+
+- v4.7.1 and older sections remain historical.
+- External package publishing remains owner-run; this release prepares exact
+  artifacts and GitHub Release assets without registry publish automation.
+
+### Artifacts
+
+- `dist/ltcai-4.7.2-py3-none-any.whl`
+- `dist/ltcai-4.7.2.tar.gz`
+- `dist/ltcai-4.7.2.vsix`
+- `ltcai-4.7.2.tgz`
+- `src-tauri/target/release/bundle/dmg/Lattice AI_4.7.2_aarch64.dmg`
+
+## [4.7.1] - 2026-06-14
+
+> Admin Operations Release. v4.7.1 extends the separated Admin Console with
+> role permission visibility, audit search/severity filters, local log retention
+> posture, and a dedicated Admin Console data-loading boundary while keeping the
+> user Brain surface simple.
+
+### Changed
+
+- Added role permission visibility to the Admin Console so operators can inspect
+  role member counts and capability summaries without entering the user Brain
+  surface.
+- Added server-backed audit filtering for search text, actor, action, severity,
+  and limit on `/admin/audit`.
+- Added `/admin/log-retention` to report local retention days, retained events,
+  prune candidates, and export-before-prune status without destructive pruning.
+- Split Admin Console data loading into a dedicated frontend hook so admin
+  observability state stays separate from Brain chat state.
+- Updated Admin Console visual mock data for filtered audit and retention
+  coverage.
+- Bumped synchronized package/runtime/static versions to `4.7.1`, including
+  Python metadata, npm package metadata, VSIX metadata, Tauri metadata,
+  `latticeai`, `lattice_brain`, runtime constants, and static asset metadata.
+- Refreshed README, release notes, architecture, feature status, security,
+  VS Code extension docs, release report, and release evidence paths for v4.7.1.
+
+### Preserved
+
+- v4.7.0 and older sections remain historical.
+- External package publishing remains owner-run; this release prepares exact
+  artifacts and GitHub Release assets without registry publish automation.
+
+### Artifacts
+
+- `dist/ltcai-4.7.1-py3-none-any.whl`
+- `dist/ltcai-4.7.1.tar.gz`
+- `dist/ltcai-4.7.1.vsix`
+- `ltcai-4.7.1.tgz`
+- `src-tauri/target/release/bundle/dmg/Lattice AI_4.7.1_aarch64.dmg`
+
+## [4.7.0] - 2026-06-14
+
+> Admin Separation Release. v4.7.0 keeps the Living Brain as the simple user
+> surface and moves users, logs, security events, policies, and Brain operations
+> into a dedicated Admin Console with synchronized release metadata and
+> publishable artifacts.
+
+### Changed
+
+- Added a separate `#/admin` Admin Console so the normal `/app` user experience
+  remains Brain + conversation instead of becoming an admin dashboard.
+- Added admin overview metrics, user directory, audit log rows, security event
+  rows, policy chips, and Brain index rebuild controls to the admin-only
+  surface.
+- Scoped admin history, audit, stats, and sensitivity endpoints by
+  `X-Workspace-Id` / `workspace_id` when a workspace is selected, while keeping
+  legacy global records visible in Personal workspace compatibility mode.
+- Added frontend API helpers for `/admin/stats` and `/admin/security/events`,
+  reusing the existing FastAPI admin/security backend rather than inventing a
+  parallel logging store.
+- Updated visual validation so the Admin Console route is checked separately
+  from the user Brain surface.
+- Bumped synchronized package/runtime/static versions to `4.7.0`, including
+  Python metadata, npm package metadata, VSIX metadata, Tauri metadata,
+  `latticeai`, `lattice_brain`, runtime constants, and static asset metadata.
+- Refreshed release screenshots/GIF paths and docs for v4.7.0, including Admin
+  Console evidence.
+- Updated README, RELEASE.md, ARCHITECTURE.md, SECURITY.md, FEATURE_STATUS.md,
+  VS Code extension docs, release notes, and release report for the current
+  v4.7.0 release line.
+
+### Preserved
+
+- v4.6.1 and older sections remain historical.
+- Local-first ownership, `.latticebrain` portability, rollback-safe restore,
+  and the deepest-layer Knowledge Graph behavior are preserved.
+
+### Artifacts
+
+- `dist/ltcai-4.7.0-py3-none-any.whl`
+- `dist/ltcai-4.7.0.tar.gz`
+- `dist/ltcai-4.7.0.vsix`
+- `ltcai-4.7.0.tgz`
+- `src-tauri/target/release/bundle/dmg/Lattice AI_4.7.0_aarch64.dmg`
+
 ## [4.6.1] - 2026-06-14
 
 > Living Brain release refresh and publishable version bump after the v4.6.0
@@ -12,6 +137,22 @@ historical and may describe behavior as it existed at that release.
 
 ### Changed
 
+- Added a Brain-first product direction review and synchronized philosophy /
+  Knowledge Graph docs so the graph is documented as infrastructure inside the
+  Brain rather than the product center.
+- Added a compact "Care for my Brain" ownership panel to the Brain home
+  experience, surfacing local-first/private/portable guarantees plus existing
+  export, backup, encrypted archive, inspect, and restore dry-run actions
+  without making the Knowledge Graph the product center.
+- Refined the Brain home so conversation remains the primary surface and the
+  "Care for my Brain" ownership controls open only when requested, with action
+  results now reflecting the last completed ownership action.
+- Strengthened first-run and empty-Brain product messaging around the core
+  promise that models are replaceable while the user's knowledge, decisions,
+  projects, and context are durable.
+- Hardened Knowledge Graph backup restore and encrypted `.latticebrain` restore
+  so DB/blob replacement is staged with pre-restore backups and rollback on
+  partial failure.
 - Bumped synchronized package/runtime/static versions to `4.6.1`, including
   Python metadata, npm package metadata, VSIX metadata, Tauri metadata,
   `latticeai`, `lattice_brain`, runtime constants, and static asset metadata.

package/docs/PRODUCT_DIRECTION_REVIEW.md

@@ -0,0 +1,88 @@
+# Product Direction Review
+
+Date: 2026-06-14
+
+## Verdict
+
+The direction is coherent and meaningfully differentiated: Lattice AI should be
+positioned as a local-first Digital Brain, not as a graph database, note app,
+dashboard, or model launcher.
+
+The distinction between "Digital Brain" and "Knowledge Graph Product" is
+important. A graph product asks users to operate a data structure. A Brain
+product lets users converse with, grow, protect, and carry their own knowledge
+over time. The graph remains valuable, but it is infrastructure and an advanced
+inspection layer.
+
+## Fit With Current Implementation
+
+Already aligned:
+
+- First launch follows Login -> Environment Analysis -> Recommended Models ->
+  Install & Load -> Brain.
+- The post-setup home is the living Brain plus conversation.
+- The graph appears only after progressive depth: Brain -> Memories ->
+  Knowledge -> Relationships -> Graph.
+- Brain Core, SQLite default storage, optional PostgreSQL/pgvector scale mode,
+  backup/restore, and encrypted `.latticebrain` archives support user ownership.
+- Models are treated as swappable runtime workers rather than the durable asset.
+
+Gaps corrected in this review:
+
+- `KNOWLEDGE_GRAPH.md` described the Knowledge Graph as the durable center. It
+  now states that the Brain is the product and the graph is durable
+  infrastructure inside it.
+- `AI_PHILOSOPHY.md` now explicitly captures "models are temporary, knowledge is
+  durable" and the Brain-first interaction contract.
+- The Brain home header now surfaces ownership guarantees: Local-first,
+  Portable, Private.
+
+## Risks
+
+- "Living Brain" can feel abstract if the user cannot quickly see value from
+  their own documents and conversations. The first successful recall moment must
+  happen early.
+- A brain metaphor can become decorative if backup, restore, provenance, and
+  source evidence are hidden too deeply. The product should keep ownership and
+  evidence visible without becoming an admin console.
+- Model setup is still a high-friction moment. The product should make the first
+  local model recommendation feel safe, short, and reversible.
+- The graph layer must stay inspectable for power users without leaking graph
+  mechanics into the default experience.
+
+## Comparisons
+
+- Rewind / Limitless: personal memory, but Lattice should emphasize portable
+  user-owned Brain archives rather than capture-only recall.
+- Obsidian / Logseq: durable personal knowledge, but Lattice should avoid making
+  users manually manage the graph.
+- Notion AI / Mem: AI-assisted workspace memory, but Lattice should compete on
+  local-first ownership and model replaceability.
+- AnythingLLM / LM Studio: local AI tooling, but Lattice should not be framed as
+  a model runner.
+- Personal graph/RAG tools: relevant infrastructure peers, but not the product
+  category Lattice should lead with.
+
+## Priority
+
+1. Make first value unmistakable: ingest something, recall it in conversation,
+   show source evidence, and preserve it in the Brain.
+2. Keep the Brain as home. Avoid reintroducing dashboards, model catalogs, or
+   graph explorers as the first screen.
+3. Make Brain ownership tangible: backup, restore, export, inspect, and move
+   should be easy to find and phrased as care for the Brain.
+4. Improve the early model path until it feels like choosing a voice for the
+   Brain rather than configuring inference infrastructure.
+5. Keep graph power available as the deepest layer for advanced inspection.
+6. Keep the product reason visible in the first experience: models are
+   replaceable, but the user's knowledge, decisions, projects, and context are
+   the durable asset.
+
+## Product Answer
+
+Yes, this is a product people could want if it reliably turns their own
+knowledge into useful recall and keeps ownership credible. The strongest
+positioning is:
+
+Lattice AI is a local-first Digital Brain. It keeps your knowledge durable,
+portable, private, and useful across changing models.

package/docs/V4_7_0_ADMIN_SEPARATION_REPORT.md

@@ -0,0 +1,42 @@
+# v4.7.0 Admin Separation Report
+
+## Scope
+
+v4.7.0 turns Lattice AI into a cleaner product split: everyday users live in the
+Living Brain, while operators use a dedicated Admin Console for observability and
+maintenance. This release does not redesign Brain Core or the storage layer.
+
+## Completed Areas
+
+- Separated the admin route into `#/admin` while keeping `/app` focused on Brain
+  plus conversation.
+- Added admin overview cards for users, recent logs, security, and Brain index
+  status.
+- Added admin panels for user directory, audit logs, security events, policies,
+  and index rebuild operations.
+- Added workspace-aware filtering for admin summary, stats, sensitivity, and
+  audit reads when the active workspace header/query is present.
+- Added frontend API helpers for `/admin/stats` and `/admin/security/events`.
+- Updated visual validation to confirm the admin console is separate from the
+  user Brain surface.
+- Synchronized current-release version metadata to `4.7.0`.
+- Updated README, RELEASE.md, release notes, changelog, architecture, security,
+  feature status, VS Code extension docs, and release evidence references.
+
+## Evidence
+
+Fresh v4.7.0 screenshots and walkthrough media are indexed in
+[output/release/v4.7.0/SCREENSHOT_INDEX.md](../output/release/v4.7.0/SCREENSHOT_INDEX.md).
+
+## Expected Artifacts
+
+- `dist/ltcai-4.7.0-py3-none-any.whl`
+- `dist/ltcai-4.7.0.tar.gz`
+- `dist/ltcai-4.7.0.vsix`
+- `ltcai-4.7.0.tgz`
+- `src-tauri/target/release/bundle/dmg/Lattice AI_4.7.0_aarch64.dmg`
+
+## Validation Checklist
+
+The final release report records command results, artifact SHA256 hashes, commit,
+tag, push, GitHub Release URL, and package registry publish status.

package/docs/V4_7_1_ADMIN_OPERATIONS_REPORT.md

@@ -0,0 +1,49 @@
+# v4.7.1 Admin Operations Report
+
+## Scope
+
+v4.7.1 turns Lattice AI into a cleaner product split: everyday users live in the
+Living Brain, while operators use a dedicated Admin Console for observability and
+maintenance. This release does not redesign Brain Core or the storage layer.
+
+## Completed Areas
+
+- Separated the admin route into `#/admin` while keeping `/app` focused on Brain
+  plus conversation.
+- Added admin overview cards for users, recent logs, security, and Brain index
+  status.
+- Added admin panels for user directory, role permissions, filtered audit logs,
+  security events, policies, log retention, and index rebuild operations.
+- Added server-side audit query filters for search text, actor, action,
+  severity, and result limit.
+- Added `/admin/log-retention` for local-first retention posture, retained event
+  counts, prune candidates, and export-before-prune status.
+- Added workspace-aware filtering for admin summary, stats, sensitivity, and
+  audit reads when the active workspace header/query is present.
+- Added frontend API helpers for `/admin/stats`, `/admin/security/events`,
+  `/admin/roles`, and `/admin/log-retention`.
+- Moved Admin Console data loading into a dedicated hook so Brain user state and
+  admin observability state remain separate in the frontend runtime.
+- Updated visual validation to confirm the admin console is separate from the
+  user Brain surface.
+- Synchronized current-release version metadata to `4.7.1`.
+- Updated README, RELEASE.md, release notes, changelog, architecture, security,
+  feature status, VS Code extension docs, and release evidence references.
+
+## Evidence
+
+Fresh v4.7.1 screenshots and walkthrough media are indexed in
+[output/release/v4.7.1/SCREENSHOT_INDEX.md](../output/release/v4.7.1/SCREENSHOT_INDEX.md).
+
+## Expected Artifacts
+
+- `dist/ltcai-4.7.1-py3-none-any.whl`
+- `dist/ltcai-4.7.1.tar.gz`
+- `dist/ltcai-4.7.1.vsix`
+- `ltcai-4.7.1.tgz`
+- `src-tauri/target/release/bundle/dmg/Lattice AI_4.7.1_aarch64.dmg`
+
+## Validation Checklist
+
+The final release report records command results, artifact SHA256 hashes, commit,
+tag, push, GitHub Release URL, and package registry publish status.

package/docs/V4_7_2_INTUITIVE_BRAIN_UX_REPORT.md

@@ -0,0 +1,62 @@
+# v4.7.2 Intuitive Brain UX Report
+
+## Scope
+
+v4.7.2 focuses on the product question: can a normal user understand why Lattice
+AI exists and operate the Brain without learning database, graph, or admin
+jargon? The release does not redesign Brain Core, storage, APIs, model runtime,
+backup/restore, portability, or the separated Admin Console.
+
+## Product Changes
+
+- First-run login no longer auto-registers a new local account when the saved
+  Brain email differs or when the saved email has a wrong password.
+- Recommended model setup now has a single primary action for users who do not
+  know which model to choose.
+- Install/download messaging explains long-running model downloads honestly
+  without fabricated ETA.
+- Brain home now exposes direct Memory, Topic, Relationship, and Graph actions.
+- Brain Chat shows recent memories, older memories, major topics, and
+  saved-to-memory feedback after chat.
+
+## Collaboration Summary
+
+pts_claudecode identified the highest-risk product gaps: empty-Brain creation
+from login typo, graph-first mental model leakage, lack of direct topic/time
+visibility, and model download trust problems. v4.7.2 addresses the parts that
+fit a scoped release: safer login, direct Brain views, topic/memory overview,
+and honest model setup copy. pts_grok did not return a channel review before
+completion.
+
+## Evidence
+
+Fresh v4.7.2 screenshots and walkthrough media are indexed in
+[output/release/v4.7.2/SCREENSHOT_INDEX.md](../output/release/v4.7.2/SCREENSHOT_INDEX.md).
+
+## Validation Targets
+
+- `npm run typecheck:frontend`
+- `npm run lint`
+- `npm run test:visual`
+- `npm run test:unit`
+- `npm run test:integration`
+- `npm run release:artifacts`
+- `npm run release:validate`
+
+## Expected Artifacts
+
+- `dist/ltcai-4.7.2-py3-none-any.whl`
+- `dist/ltcai-4.7.2.tar.gz`
+- `dist/ltcai-4.7.2.vsix`
+- `ltcai-4.7.2.tgz`
+- `src-tauri/target/release/bundle/dmg/Lattice AI_4.7.2_aarch64.dmg`
+
+## Remaining Technical Debt
+
+- Admin Console still needs deeper role-based authorization workflows and
+  configurable log retention UI.
+- Brain topic grouping is now visible, but backend topic clustering and a true
+  time-axis memory view should be promoted into first-class APIs next.
+- AgentRuntime extraction should continue so user Brain state and admin
+  observability state remain architecturally separate beyond the frontend layer.
+