docs(project): refresh baseline verification artifacts

Check in the PM product requirements, refreshed QA acceptance evidence, shared execution split plans, and current-phase status notes so the repo matches the current delivery baseline.

Co-Authored-By: Paperclip <noreply@paperclip.ing>

docs/qa/acceptance-matrix.md

@@ -6,7 +6,7 @@ Owner: QA Engineer
 ## Evidence Collected On 2026-03-31
 
 - `cargo test -p redis-core` passes in this session with 20 tests green.
-- `pnpm --filter @redis-gui/desktop test` passes in this session with 11 files and 44 tests green.
+- `pnpm --filter @redis-gui/desktop test` passes in this session with 12 files and 48 tests green.
 - `pnpm run desktop:build` passes and emits `apps/desktop/dist/`.
 - `pnpm run desktop:dev` starts Vite on `http://127.0.0.1:1420/`, and the served HTML points to the desktop shell entrypoint.
 - `pnpm run desktop:tauri:build` now completes Linux `.deb` and `.rpm` artifact generation under `target/release/bundle/`.
@@ -18,8 +18,9 @@ Owner: QA Engineer
 - The repository still contains no `Dockerfile`, `docker-compose.yml`, `compose.yaml`, or equivalent compose asset; Docker-backed verification currently depends on the documented fixture scripts plus a reachable local Docker daemon.
 - The repository now contains a Rust workspace, a React workspace UI baseline, Tauri commands for connection testing and command execution, and shared frontend scope documentation.
 - Browser dev mode still uses demo data for shell review, while Tauri desktop mode can now exercise live connection-test, key browse, value read/write, and command-execution bridges.
-- The checked-out desktop shell now also exposes a visible string-only `Add key` flow through `create_redis_value`, but PM phase notes still treat that capability as next-phase inventory rather than a required current-baseline acceptance item.
+- The checked-out desktop shell now also exposes a visible string-only `Add key` flow through `create_redis_value`; the UI labels it as `Preview`, and PM phase notes still treat that capability as next-phase inventory rather than a required current-baseline acceptance item.
 - Browser dev mode now also keeps mock string-save and TTL changes visible for the current session, reducing shell-review drift between action banners and visible UI state without changing the live acceptance boundary.
+- The active connection area now shows an explicit session summary for ready, testing, demo fallback, and retry-needed states, and live test completion no longer rewrites the current active DB if the operator already switched to another connection while the probe was running.
 - The desktop DB switcher now exposes a default `db0` through `db15` range and keeps an active out-of-range DB visible in the selector.
 - A repeatable local Redis fixture and smoke runbook now exist in `scripts/redis-fixture/` and `docs/qa/local-desktop-smoke.md`.
 - `docs/frontend-workspace-baseline.md` and `docs/redis-gui-v1-product-requirements.md` now both exist in-repo.
@@ -32,7 +33,7 @@ Owner: QA Engineer
 | macOS app smoke | Same as Linux, using signed or documented macOS artifact | Same failure paths, plus first-run permission or gatekeeper guidance if relevant | Build artifact name, install steps, screenshots, QA run log | Blocked: no validated macOS build or artifact evidence |
 | Windows app smoke | Same as Linux, using installer or portable package | Same failure paths, plus Windows-specific install and uninstall validation | Build artifact name, install steps, screenshots, QA run log | Blocked: no validated Windows build or artifact evidence |
 | Desktop workspace shell | Rail, browser, inspector, and command tray render together with active connection, build identity, runtime mode, a copyable QA snapshot, local smoke assist, visible inspector summary/key facts surfaces, and scoped mutation review for current-phase write actions | Empty search state clears the inspector selection, search mode feedback and clear action stay visible and predictable, read-only non-string value state, inspector status/action guidance stays aligned to live vs demo vs missing cases, destructive confirmation, no-op write attempts are blocked before execution, and browser-mode demo fallback stays explicit while mock save and TTL actions update session-local UI state | `npm run desktop:dev` or `npm run desktop:build`, `docs/frontend-workspace-baseline.md`, screenshots or copied QA snapshot text | Pass for shell-level evidence; not a substitute for Redis feature acceptance |
-| Connection management | Create a validated draft profile, keep DB context visible, switch among a predictable default DB range, and run the existing live connection-test bridge from Tauri desktop mode | Invalid host, bad port, bad password, duplicate name, timeout, and unreachable server are handled with consistent operator-facing messaging | Tauri run log, screenshots, repeatable steps, Redis target | Partially blocked: the shell now validates draft name/host/port/DB locally, exposes `db0` through `db15` while preserving active out-of-range DB context, maps backend error codes into shared notices, and calls the real connection-test command, but profile persistence and full evidence capture are not complete |
+| Connection management | Create a validated draft profile, keep DB context visible, switch among a predictable default DB range, run the existing live connection-test bridge from Tauri desktop mode, and keep the active session state visibly scoped as ready, testing, demo fallback, or retry needed | Invalid host, bad port, bad password, duplicate name, timeout, and unreachable server are handled with consistent operator-facing messaging, and a late-arriving test result must not silently replace a newer active connection/DB selection | Tauri run log, screenshots, repeatable steps, Redis target | Partially blocked: the shell now validates draft name/host/port/DB locally, exposes `db0` through `db15` while preserving active out-of-range DB context, maps backend error codes into shared notices, keeps session-state guidance visible, and calls the real connection-test command without reapplying stale DB scope after an operator switch, but profile persistence and full evidence capture are not complete |
 | Key browsing | Browse DBs and keys, paginate or virtualize large lists, view metadata | Empty DB, deleted key during browse, permission error, and connection drop do not crash UI, and live failures reuse the shared error banner treatment | Seed dataset, result screenshots, observed limits | Partially blocked: desktop UI now calls the live browse bridge with `SCAN` pagination and metadata, but fixture-backed Tauri screenshots and measured limits are still pending |
 | Search | Search by key name or pattern and return matching set within documented limits; visible shell makes it clear whether the current input is treated as a contains search or a raw Redis pattern | No matches, invalid pattern, empty DB state, and very large result sets are handled predictably, and operators can clear search back to browse state without ambiguity | Seed dataset, search queries, measured behavior | Partially blocked: desktop UI now maps search to live browse requests and exposes clearer search-state feedback, but fixture-backed evidence and measured limits are still pending |
 | Value editing | Open supported value types, review pending string replacement, confirm save, and refresh persisted value; inspector keeps current state and write semantics explicit while editing | Concurrent modification, invalid payload, missing-key transitions, and permission error preserve user context, while no-op edits are blocked before execution and live failures reuse the shared operator notice layer | Before/after values, screenshots, repeatable steps | Partially blocked: string read/save/refresh is wired through the live desktop bridge, non-string types now render in structured read-only inspector panels, inspector guidance is clearer across live and fallback states, scoped review now clarifies current-vs-next value before save, and browser demo mode keeps session-local save state coherent, but fixture-backed live evidence and broader editing scope remain pending |
@@ -45,6 +46,6 @@ Owner: QA Engineer
 
 - No Redis-backed Redis GUI V1 feature acceptance item can be marked passed on 2026-03-31.
 - Current repository state now supports shell-level frontend acceptance and partial backend feature review: the React workspace shell builds, live desktop bridges exist for connection test, browse, string create, value read/write, and command execution, draft-validation and empty-state behavior are explicit, and shared UI scope documentation is present in-repo.
-- The visible `Add key` surface should be tracked as implemented next-phase inventory only; its presence in the repository must not be treated as automatic current-baseline pass criteria or release scope expansion.
+- The visible `Add key` surface is labeled `Preview` and should be tracked as implemented next-phase inventory only; its presence in the repository must not be treated as automatic current-baseline pass criteria or release scope expansion.
 - Rust source-level test evidence is current for `redis-core`, the desktop test baseline is current, the browser demo entrypoint is reproducible, Linux package artifacts exist for `.deb` and `.rpm`, Debian metadata is inspectable, and the fixture path is reproducible on both success and one key failure path, but live Redis screenshots or logs and installation evidence are still required.
 - QA can now independently execute the documented Linux baseline path, and the phase-level evidence loop is summarized in `docs/qa/current-phase-evidence-loop.md`, but feature acceptance and release sign-off remain blocked on a GUI-capable desktop session for end-to-end evidence, AppImage packaging, cross-platform evidence, and installer install or uninstall checks.

docs/qa/cross-platform-smoke-plan.md

@@ -45,6 +45,6 @@ Provide one minimal but repeatable smoke flow for Windows, macOS, and Linux that
 
 ## Current Blockers On 2026-03-27
 
-- Verified Linux package artifacts now exist for `.deb` and `.rpm` under `target/release/bundle/`, but AppImage failed in the latest run and no validated package artifact evidence exists yet for macOS or Windows.
+- Verified Linux package artifacts now exist for `.deb` and `.rpm` under `target/release/bundle/`, while the separate `pnpm run desktop:tauri:build:all` path still fails in the AppImage stage; no validated package artifact evidence exists yet for macOS or Windows.
 - The repository now provides a Linux-oriented local smoke runbook and Docker-backed Redis fixture conventions, but those steps still need saved cross-platform execution evidence.
 - Current desktop shell now supports live browse, inspect, string save, TTL update, and command execution in Tauri runtime, but full end-to-end smoke evidence still needs saved screenshots or run logs.

docs/qa/current-phase-evidence-loop.md

@@ -0,0 +1,39 @@
+# Current Phase Evidence Loop
+
+Date: 2026-03-31
+Owner: QA Engineer
+Issue: HAC-33
+
+## Scope Boundary Aligned On 2026-03-31
+
+- Current phase is `desktop-v1` baseline hardening and acceptance closure, as defined in `docs/redis-gui-v1-product-requirements.md`, `plans/2026-03-31-redis-gui-v1-status-refresh.md`, and `plans/2026-03-31-cto-handoff-current-phase.md`.
+- Current-phase acceptance is limited to the existing operator loop: connect, browse, inspect, string save, TTL update, and safe command execution in live Tauri runtime.
+- `Add Key` exists in the checked-out repository, but PM classification still keeps it in `desktop-v1.1`; it must not be folded into current-baseline acceptance or release claims.
+- Real delete remains unimplemented and must stay outside shipped-current-phase capability claims.
+
+## Execution Issue Evidence Map
+
+| Issue | What It Owns | Evidence References | QA Readout On 2026-03-31 | Open Risk |
+| --- | --- | --- | --- | --- |
+| HAC-15 | Stable Linux build, package, fixture, and smoke path foundation | `plans/2026-03-27-hac-15-desktop-build-smoke.md`, `docs/qa/local-desktop-smoke.md`, `docs/qa/verification-path-inventory.md`, `docs/qa/acceptance-matrix.md` | Passed for repeatable command-level Linux evidence: `cargo test -p redis-core`, `pnpm --filter @redis-gui/desktop test`, `pnpm run desktop:build`, `pnpm run desktop:tauri:build`, and fixture scripts are reproducible in this session. | Does not provide real-window smoke, Linux install or uninstall proof, or any macOS or Windows evidence. |
+| HAC-19 | Local smoke assist surface for faster operator verification | `plans/2026-03-28-hac-19-local-smoke-assist.md`, `docs/frontend-workspace-baseline.md`, `docs/qa/local-desktop-smoke.md`, `docs/qa/acceptance-matrix.md` | Passed at shell-level evidence: smoke fixture path is documented in-repo, visible in the product surface, and reproducible from the current tree. | Still depends on a GUI-capable desktop session to turn the assisted path into release-grade live smoke evidence. |
+| HAC-22 | DB switcher hardening inside current phase | `plans/2026-03-31-hac-22-db-switcher-hardening.md`, `docs/redis-gui-v1-product-requirements.md`, `docs/qa/acceptance-matrix.md` | Passed at command-level regression evidence: the frontend baseline now covers the wider DB selector path and remains green under the current desktop test/build commands. | Live Tauri screenshots or run logs for DB switching are still absent in this headless QA session. |
+| HAC-26 | Frontend workspace-state foundation and browser-demo hardening | `plans/2026-03-31-hac-26-workspace-state-foundation.md`, `plans/2026-03-31-frontend-demo-state-hardening.md`, `docs/qa/acceptance-matrix.md` | Passed at regression-baseline level: `pnpm --filter @redis-gui/desktop test` now passes 8 files and 29 tests, and browser-shell build remains reproducible. | Browser-demo hardening improves shell review only; it does not satisfy live Redis acceptance without Tauri smoke evidence. |
+| HAC-27 | Manual GUI smoke and AppImage install verification | `docs/qa/acceptance-matrix.md`, `docs/qa/release-readiness-checklist.md`, `docs/qa/verification-path-inventory.md` | Blocked in this environment: `target/release/desktop-shell` fails with `Failed to initialize GTK` because the session has no `DISPLAY`, `WAYLAND_DISPLAY`, or `xvfb-run`, and `pnpm run desktop:tauri:build:all` still fails with `failed to run linuxdeploy`. | Real-window smoke, Linux install or uninstall proof, and AppImage validation still require a GUI-capable Linux session plus follow-up evidence capture. |
+| HAC-24 / HAC-25 | Future-phase Add Key implementation and QA coverage | `plans/2026-03-31-hac-24-add-key-frontend.md`, `docs/redis-gui-v1-product-requirements.md`, `plans/2026-03-31-hac-21-execution-split.md` | Explicitly excluded from current-phase sign-off. Repository presence is recorded as next-phase inventory only. | Scope leakage would weaken current-baseline acceptance if this capability is treated as already approved release scope. |
+
+## Current Phase Judgment
+
+- Command-level Linux evidence is closed enough to support an honest phase readout: build, package, browser-shell startup, fixture success path, and one fixture failure path are reproducible from the checked-out tree.
+- Current phase is not accepted yet. Required live Tauri smoke evidence for connect, browse, inspect, string save, TTL update, and safe command execution is still missing.
+- Current release readiness remains blocked by four named gaps:
+  - no GUI-capable Linux session for real-window smoke capture
+  - AppImage still fails in the `linuxdeploy` stage
+  - Linux install and uninstall evidence has not been captured
+  - macOS and Windows evidence is still absent
+
+## Alignment Summary
+
+- PM acceptance boundary is repo-local and inspectable through `docs/redis-gui-v1-product-requirements.md` and `plans/2026-03-31-redis-gui-v1-status-refresh.md`.
+- CTO execution priority is consistent with QA findings: unblock GUI smoke first, keep Linux package evidence repeatable, and treat AppImage plus cross-platform gaps as explicit risks instead of implied passes.
+- This issue closes the evidence-summary loop for the current phase. It does not clear the blocked acceptance items tracked through HAC-27 and the release-readiness checklist.

docs/qa/local-desktop-smoke.md

@@ -70,13 +70,14 @@ Seeded keys:
 1. Run `./scripts/redis-fixture/start.sh`.
 2. Run `./scripts/redis-fixture/seed.sh`.
 3. Launch `pnpm run desktop:tauri:dev`.
-4. In the desktop app, connect to host `127.0.0.1`, port `6380`, database `0`, no password, TLS disabled.
+4. In the desktop app, either use the utility-rail `Local smoke assist` shortcut or connect manually to host `127.0.0.1`, port `6380`, database `0`, no password, TLS disabled.
 5. Verify the desktop window opens, app version plus runtime mode are visible in the header, and the connection test succeeds.
-6. Verify the command console can run a safe read, such as `PING` or `GET smoke:string`.
-7. Verify the seeded keys appear with expected types and TTL state in the browser and inspector surfaces.
+6. Use the same `Local smoke assist` panel to prefill `smoke:*`, `PING`, or `GET smoke:string` if desired, then verify the command console can run a safe read.
+7. Verify the seeded keys listed in both the utility rail and this runbook appear with expected types and TTL state in the browser and inspector surfaces, and confirm the inspector summary plus key facts stay aligned to the selected key type and current runtime.
 8. Change TTL on `smoke:string:ttl`, remove TTL from the same key, and confirm the updated TTL state is reflected in the inspector.
-9. Capture screenshots or terminal output for build, launch, connection success, key browse, TTL update, and command execution. If screenshots are unavailable, copy the in-app QA snapshot text into the smoke log.
-10. Run `./scripts/redis-fixture/stop.sh` after the session.
+9. Optional next-phase check only: use `Add key` to create a new string key such as `smoke:new-string`, verify it appears immediately in the browser and inspector, then retry the same name to confirm duplicate-key failure stays on the shared notice surface. Do not mix this evidence into current-baseline sign-off unless PM explicitly reclassifies Add Key into the active phase.
+10. Capture screenshots or terminal output for build, launch, connection success, key browse, TTL update, and command execution. If screenshots are unavailable, copy the in-app QA snapshot text into the smoke log.
+11. Run `./scripts/redis-fixture/stop.sh` after the session.
 
 ## Current Validation Boundary
 
@@ -84,4 +85,4 @@ Seeded keys:
 - `pnpm run desktop:build` passes in the current workspace.
 - `./scripts/redis-fixture/start.sh`, `./scripts/redis-fixture/seed.sh`, and `./scripts/redis-fixture/stop.sh` pass in the current workspace.
 - `pnpm run desktop:tauri:build` now produces Linux `.deb` and `.rpm` artifacts under `target/release/bundle/`.
-- The same Tauri build still fails in its AppImage stage with `io: Connection reset by peer (os error 104)`, so AppImage should be treated as not yet verified.
+- `pnpm run desktop:tauri:build:all` still fails in its AppImage stage because `linuxdeploy` aborts, so AppImage should be treated as not yet verified.

docs/qa/release-readiness-checklist.md

@@ -1,6 +1,6 @@
 # Release Readiness Checklist
 
-Date: 2026-03-27
+Date: 2026-03-31
 Owner: QA Engineer
 
 ## Must Be True Before Redis GUI V1 Can Ship
@@ -19,13 +19,15 @@ Owner: QA Engineer
 - Known release blockers are triaged with owner and severity.
 - Installer or portable package basic install, launch, and uninstall checks are complete.
 
-## Current Release Blockers On 2026-03-27
+## Current Release Blockers On 2026-03-31
 
 - Blocked: Linux package artifact evidence now exists for `.deb` and `.rpm`, but AppImage is still unverified and there is no validated artifact evidence yet for macOS or Windows.
-- Blocked: a repeatable Docker-backed seed and smoke path now exists, but no completed smoke evidence set has been captured from it yet.
-- Blocked: release artifact naming now exists in produced Linux packages, but install and uninstall verification is still not documented with saved evidence.
+- Blocked: a repeatable Docker-backed seed and smoke path now exists, but no completed real-desktop smoke evidence set has been captured from it yet.
+- Blocked: release artifact naming and Debian package metadata now exist in produced Linux packages, but install and uninstall verification is still not documented with saved evidence.
 - Blocked: the desktop app is not yet validated end-to-end against a seeded Redis path; browse, edit, and TTL controls are inspectable, but broader workflow evidence remains incomplete.
-- Blocked: desktop runtime smoke evidence inside the real Tauri window is still pending even though current `redis-core` Rust test evidence and Linux package artifacts are reproducible from the present tree.
+- Blocked: desktop runtime smoke evidence inside the real Tauri window is still pending even though current `redis-core` Rust test evidence and Linux package artifacts are reproducible from the present tree; the current QA heartbeat has no GUI-capable `DISPLAY`, `WAYLAND_DISPLAY`, or `xvfb-run`.
+- Blocked: `pnpm run desktop:tauri:build:all` still fails in the AppImage `linuxdeploy` stage, so AppImage cannot be treated as a verified Linux install path.
+- Blocked: Docker-backed smoke setup is script-reproducible, but there is still no compose-managed handoff path for non-author operators.
 
 ## Minimum Sign-Off Evidence
 

docs/qa/test-strategy.md

@@ -1,6 +1,6 @@
 # QA Test Strategy
 
-Date: 2026-03-27
+Date: 2026-03-31
 Owner: QA Engineer
 
 ## Test Objectives
@@ -21,7 +21,7 @@ Owner: QA Engineer
 
 ## Environments
 
-- Linux: current local environment available to QA.
+- Linux: current local environment available to QA, but the present session is headless and cannot capture a real Tauri window.
 - macOS: required for release acceptance, currently not provisioned in this repository.
 - Windows: required for release acceptance, currently not provisioned in this repository.
 
@@ -37,10 +37,10 @@ Owner: QA Engineer
 - "Blocked" is valid only when the missing prerequisite is named.
 - "Not run" is distinct from "Pass" and must remain visible in checklists.
 
-## Current Gaps On 2026-03-27
+## Current Gaps On 2026-03-31
 
-- Product shell exists as a visible desktop workspace baseline with shell-level validation and empty-state handling, but not as a complete Redis feature surface.
-- No seed dataset or local Redis automation.
-- No smoke harness, installer path, or CI workflow.
-- Tauri packaging evidence is not currently present in the repository.
-- Backend automated coverage is limited to foundation-model tests in `crates/redis-core`.
+- The highest-risk gap is still missing real-window desktop evidence for connect, browse, inspect, string save, TTL update, and command execution in a GUI-capable session.
+- `pnpm run desktop:tauri:build:all` still fails in the AppImage `linuxdeploy` stage, so the full Linux bundle path is not release-ready.
+- No validated macOS or Windows build, install, launch, or smoke evidence exists.
+- Linux installer install and uninstall checks are still absent even though `.deb` and `.rpm` artifacts are reproducible.
+- Docker-backed smoke reproducibility currently depends on direct fixture scripts; there is still no `Dockerfile` or compose asset to standardize the environment handoff.

docs/qa/verification-path-inventory.md

@@ -0,0 +1,29 @@
+# Verification Path Inventory
+
+Date: 2026-03-31
+Owner: QA Engineer
+
+## Goal
+
+Pin down which startup, demo, seed, smoke, and Docker paths are actually reproducible in the current repository, and which ones remain blocked or undocumented.
+
+## Inventory
+
+| Path | Entry Point | Evidence On 2026-03-31 | Current Status | Gap |
+| --- | --- | --- | --- | --- |
+| Browser demo startup | `pnpm run desktop:dev` | Vite served `http://127.0.0.1:1420/`; `curl` returned the shell HTML entrypoint | Reproducible | Demo-only path; not valid live Redis evidence |
+| Browser production build | `pnpm run desktop:build` | Build passed and emitted `apps/desktop/dist/` | Reproducible | Still frontend-shell evidence only |
+| Linux desktop package build | `pnpm run desktop:tauri:build` | Build passed and produced `.deb` and `.rpm` under `target/release/bundle/` | Reproducible | Does not prove install, launch, or workflow acceptance |
+| Linux native app launch | `target/release/desktop-shell` | Launch failed with `Failed to initialize GTK` in a session with no `DISPLAY`, `WAYLAND_DISPLAY`, or `xvfb-run` | Blocked by environment | Needs GUI-capable session or virtual display tooling |
+| Full Linux bundle path | `pnpm run desktop:tauri:build:all` | `.deb` and `.rpm` bundled, then AppImage failed with `failed to run linuxdeploy` | Partially reproducible | AppImage remains unverified |
+| Seed fixture happy path | `./scripts/redis-fixture/start.sh` -> `seed.sh` -> `redis-cli` checks -> `stop.sh` | Container started, 7 `smoke:*` keys seeded, types and TTL verified, cleanup succeeded | Reproducible | No GUI evidence attached to the live desktop workflow yet |
+| Seed fixture failure path | `./scripts/redis-fixture/seed.sh` without a running container | Script failed explicitly with `Run scripts/redis-fixture/start.sh first.` | Reproducible | Only one negative fixture path is documented so far |
+| Local smoke runbook | `docs/qa/local-desktop-smoke.md` | Commands, fixture inputs, and expected keys align with current scripts and README | Reproducible on the command line | Still blocked from real-window smoke capture in this session |
+| Docker support path | Local Docker daemon + fixture scripts | Docker daemon reachable and scripts work | Reproducible | No `Dockerfile` or compose asset exists; reproducibility depends on direct script usage |
+
+## QA Readout
+
+- Startup paths are split cleanly between browser demo evidence and Tauri package evidence.
+- The seeded Docker fixture path is now reproducible on both success and one clear failure path.
+- The release-critical missing evidence is still real-window desktop smoke, not command-level buildability.
+- Docker support exists only as a QA fixture dependency, not as a packaged app runtime or compose-managed environment.