Update docs for UX feedback wave 1 delivery

- STATUS.md: record 6 merged PRs (#612, #615, #617, #621, #623, #626),
  update BoardContextBuilder description for card IDs and N+1 fix,
  update test counts to 1949 backend / 1444 frontend
- IMPLEMENTATION_MASTERPLAN.md: add UX wave 1 to premium UI stream,
  update test counts
- Tracker #628: check off 8 delivered issues

Author: Chris0Jeky

docs/IMPLEMENTATION_MASTERPLAN.md

@@ -55,8 +55,8 @@ Delivered in the latest cycle:
    - request correlation middleware + Ops CLI correlation propagation
    - timing/result diagnostics for log query and automation execution paths
 7. Test surface expanded and verified:
-   - Backend: 1668+ passing (property-based and fuzz tests added via FsCheck)
-   - Frontend unit: 1174+ passing (123+ test files; batch triage, search, accessibility tests added)
+   - Backend: 1949+ passing (property-based and fuzz tests added via FsCheck)
+   - Frontend unit: 1444+ passing (132+ test files; batch triage, search, accessibility tests added)
    - Default Playwright regression lane: 24+ passing (accessibility axe-core E2E added; `stakeholder-demo.spec.ts` remains opt-in/skipped by default)
 8. Documentation consolidation retained:
    - active docs remain focused at `docs/` root
@@ -701,7 +701,7 @@ Exit Criteria:
 These continue in parallel where they protect trust, performance, or operator posture, but they should not outrun Horizon A through C product legibility work:
 
 - managed-key LLM control plane and abuse controls: `#235`, `#237` (pending), `#238` (operator tooling groundwork delivered; live-traffic wiring pending), `#239` (delivered), `#240` (delivered)
-- premium UI foundations and reskin wave: `#242` to `#250` (plus optional `#251`); foundations delivered: `#243` UI-02 shared primitives, `#245` UI-03 stack spike, `#250` PERF-08 budgets; appshell reskin (`#499`) and board/card polish (`#501`) now shipped with design-token-based styling
+- premium UI foundations and reskin wave: `#242` to `#250` (plus optional `#251`); foundations delivered: `#243` UI-02 shared primitives, `#245` UI-03 stack spike, `#250` PERF-08 budgets; appshell reskin (`#499`) and board/card polish (`#501`) now shipped with design-token-based styling; UX feedback wave 1 (`#628`) delivered: sidebar footer pinned (`#623`), card drag layout shift eliminated (`#621`), starter-pack modal migrated to design tokens (`#612`), capture triage error messages (`#615`), review collapsible sections with risk color-coding (`#626`)
 - long-list responsiveness and related UX scale follow-through: `#213` (delivered — inbox + activity virtualized; board cards deferred due to drag-and-drop conflicts)
 - platform, ops, testing, and maturity backlog: `#84` to `#111`, `#87` to `#91`
 - deferred outreach CRM expansion: `#262` to `#268`

docs/STATUS.md

@@ -24,7 +24,7 @@ Rebranding thesis (2026-02-23):
 Current constraints are mostly hardening and consistency:
 - security and identity behavior is converging but still not uniform across all controller families
 - some UX/operator surfaces are functional but not yet keyboard-first or discoverability-first
-- LLM flow now supports config-gated `OpenAI` and `Gemini` providers with deterministic `Mock` fallback for safe local/test posture; degraded provider responses are now structurally distinct (`messageType: "degraded"` + `degradedReason`) and the health endpoint supports opt-in probe verification (`?probe=true`); chat-to-proposal pipeline improvements delivered: `LlmIntentClassifier` now uses compiled regex patterns with word-distance matching, stemming/plurals, broader verb coverage, and negative context filtering for negations and other-tool questions (`#571`); parse failures now return structured hint payloads with closest-match suggestions and a frontend hint card with "try this instead" pre-fill (`#572`); dedicated classifier and chat-to-proposal integration test coverage added (`#577`); LLM-assisted instruction extraction now delivered (`#573`): OpenAI and Gemini providers request structured JSON output with a system prompt describing supported instruction patterns, parse the response into `LlmCompletionResult.Instructions`, and fall back to the static `LlmIntentClassifier` when structured parsing fails; `ChatService` iterates LLM-extracted instructions (supporting multiple proposals from a single message) and falls back to raw user message parsing when no instructions are extracted; Mock provider unchanged for deterministic test behavior; multi-instruction batch parsing now delivered (`#574`): `ParseBatchInstructionAsync` splits multiple natural-language instructions into individual planner calls, `ChatService` routes multi-instruction messages through batch parsing to generate multiple proposals from a single chat message; board-context LLM prompting now delivered (`#575`): `BoardContextBuilder` constructs bounded board context (columns, card titles, labels) and appends it to system prompts across OpenAI and Gemini providers via `LlmSystemPromptBuilder`; **remaining gap**: conversational refinement (`#576`) remains undelivered; analysis at `docs/analysis/2026-03-29_chat_nlp_proposal_gap.md`
+- LLM flow now supports config-gated `OpenAI` and `Gemini` providers with deterministic `Mock` fallback for safe local/test posture; degraded provider responses are now structurally distinct (`messageType: "degraded"` + `degradedReason`) and the health endpoint supports opt-in probe verification (`?probe=true`); chat-to-proposal pipeline improvements delivered: `LlmIntentClassifier` now uses compiled regex patterns with word-distance matching, stemming/plurals, broader verb coverage, and negative context filtering for negations and other-tool questions (`#571`); parse failures now return structured hint payloads with closest-match suggestions and a frontend hint card with "try this instead" pre-fill (`#572`); dedicated classifier and chat-to-proposal integration test coverage added (`#577`); LLM-assisted instruction extraction now delivered (`#573`): OpenAI and Gemini providers request structured JSON output with a system prompt describing supported instruction patterns, parse the response into `LlmCompletionResult.Instructions`, and fall back to the static `LlmIntentClassifier` when structured parsing fails; `ChatService` iterates LLM-extracted instructions (supporting multiple proposals from a single message) and falls back to raw user message parsing when no instructions are extracted; Mock provider unchanged for deterministic test behavior; multi-instruction batch parsing now delivered (`#574`): `ParseBatchInstructionAsync` splits multiple natural-language instructions into individual planner calls, `ChatService` routes multi-instruction messages through batch parsing to generate multiple proposals from a single chat message; board-context LLM prompting now delivered (`#575`, expanded in `#617`): `BoardContextBuilder` constructs bounded board context (columns, card IDs, titles, labels) grouped per column and appends it to system prompts across OpenAI and Gemini providers via `LlmSystemPromptBuilder`; card IDs are included as first-8 hex chars so the LLM can generate `move card <id>` instructions; context budget increased to 4000 chars with single-query card fetch; **remaining gap**: conversational refinement (`#576`) remains undelivered; analysis at `docs/analysis/2026-03-29_chat_nlp_proposal_gap.md`
 - managed-key shared-token abuse-control strategy is now explicitly seeded in `#235` to `#240` before broad external exposure
 - testing-harness guardrail expansion from `#254` to `#260` is shipped; remaining work is normal follow-up hardening rather than the original wave
 - MVP dogfooding flow now supports canonical checklist bootstrap in chat (proposal-first, board-scoped); broader template coverage remains future work
@@ -39,6 +39,7 @@ Current constraints are mostly hardening and consistency:
 - fresh-registration manual test (2026-03-29) surfaced 2 P0 blockers and 16 additional bugs/observations spanning data isolation, board stability, dark-mode theming, chat utility, and UX polish; full findings at `docs/analysis/2026-03-29_manual_testing_consolidated_findings.md`; P0 blockers: queue data not scoped to the authenticated user (`#508`), board auto-switching on multi-board accounts (`#509`); P1 issues tracked in `#510` through `#515`; P2/P3 in `#516` through `#524`; **no external user onboarding should occur until `#508` is resolved**; two bugs from this session now resolved: activity audit trail not recording board mutations (`#521`, fixed — audit logging wired for all board/card/column/label mutations with `SafeLogAsync` resilience wrapper), archive board 30-second freeze (`#519`, fixed — navigation before reactive state teardown prevents cascading re-renders)
 - platform expansion strategy (2026-03-29) now covers four strategic pillars: market adoption (`#544`), packaging/distribution (`#532`), cloud/collaboration (`#537`), and mobile platform (`#540`); master strategy tracker at `#531`; strategy documents at `docs/strategy/`; release versioning plan: `v0.1.0` (self-contained exe) → `v0.2.0` (hosted cloud) → `v0.3.0` (PWA/mobile) → `v0.4.0` (collaboration) → `v0.5.0` (platform maturity) → `v1.0.0` (GA)
 - hands-on UX testing (2026-03-31) surfaced 8 areas of feedback spanning review, inbox, today, home, board, notifications, and LLM chat; 2 P1 issues (capture triage fails on natural-language text `#614`, chat response truncation showing raw JSON `#616`), 11 P2 usability/visual-coherence issues (`#611`–`#613`, `#615`, `#617`, `#620`, `#622`, `#624`–`#627`), and 4 P3 polish/strategic items (`#618`–`#619`, `#621`, `#623`); tracker at `#628`; full analysis at `docs/analysis/2026-03-31_manual_testing_ux_feedback.md`; cross-cutting themes: progressive disclosure needed across review/today/notifications, semantic color vocabulary for status tags, capture pipeline needs LLM-assisted extraction, and chat needs tool-calling/function-calling architecture; LLM tool-calling spike (`#618`) and MCP server spike (`#619`) seeded for strategic planning
+- UX feedback wave 1 delivered (2026-03-31): 6 of 17 issues from `#628` resolved — sidebar footer pinned (`#623`), card drag layout shift eliminated (`#621`), starter-pack modal migrated to design tokens (`#612`), capture triage error messages surfaced with retry hint (`#615`), board context expanded with card IDs for LLM chat (`#617`, N+1 query fixed), review proposal cards now use collapsible detail sections with risk color-coding and keyboard-accessible links dropdown (`#626`); remaining open: 2 P1 (`#614`, `#616`), 5 P2 (`#613`, `#620`, `#624`, `#625`, `#627`), 4 P3 (`#618`, `#619`)
 
 Target experience metrics for the capture direction:
 - capture action to saved artifact should feel under 10 seconds in normal use
@@ -710,12 +711,12 @@ Command:
 - `dotnet test backend/Taskdeck.sln -c Release -m:1`
 
 Result:
-- Domain: 306/306 passing
-- Application: 943/943 passing
-- API integration: 407/407 passing
+- Domain: 357/357 passing
+- Application: 1167/1167 passing
+- API integration: 413/413 passing
 - CLI contract: 4/4 passing
 - Architecture boundaries: 8/8 passing
-- Backend Total: 1668/1668 passing
+- Backend Total: 1949/1949 passing
 
 ### Frontend Unit + Build (Executed)
 
@@ -726,7 +727,7 @@ Commands:
 - `cd frontend/taskdeck-web && npm run build`
 
 Result:
-- Frontend unit: 1174/1174 passing (123 test files)
+- Frontend unit: 1444/1444 passing (132 test files)
 - Typecheck: passing
 - Production build: passing