Two small ergonomic fixes after the impeccable pass:
- TaskLane keyboard hints (⏎ submit · ⇧⏎ newline) under each open input
were rendered at text-muted-foreground/70, just shy of legible at a
glance. Drop the /70 opacity modifier so they read at full muted weight
on first look without becoming visually loud.
- 12 sites across the session screen had explicit font-sans utilities,
but the body default is already IBM Plex Sans (via --font-sans in
index.css and Tailwind v4's default-sans binding). None of the call
sites sit inside a font-heading or font-mono cascade, so every
font-sans there was a no-op. Drop them. ConcludeSessionModal also had
three "text-xs font-sans text-xs" triplets — drop both the redundant
font-sans and the doubled text-xs in one pass.
Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
Multi-step UX refactor of the assistant chat session screen, run via the
$impeccable skill. Heuristic score moved 24/40 → 33/40 (+9), with the biggest
gains on Aesthetic & Minimalist (1→3), Consistency & Standards (1→3), and
Recognition Rather Than Recall (2→4).
Distill — chat region:
- Remove the "Suggested checks" chip strip + selected-chip detail card; the
TaskLane is the single canonical home for "what to do next"
- Add an inline Next steps · N pending cue above the latest action-bearing
AI bubble (anchors attention without duplicating the lane's items)
- Link banner ↔ script-panel lifecycle: collapsing or dismissing the
ProposalBanner now also hides the InlineNoTemplateDialog / TemplateMatchPanel
- Drop backdrop-blur on the handoff-context overlay (DESIGN-SYSTEM hard rule)
Quieter — drop decoration overshoot:
- Remove 3px side stripes on TaskLane done cards, all 6 ProposalBanner modes,
WhatWeKnowItem fact rows
- Drop bg-gradient surfaces on WhatWeKnow + every ProposalBanner mode
- Drop 2px accent borderTop on the TaskLane header
- Replace bordered avatar boxes in banners with inline state-colored icons
- Each surface now uses a single decoration channel (top border + inline icon)
Layout:
- Header consolidates to Resolve + Escalate + ⋯ kebab; Context, New Ticket,
Update Ticket, Pause now live behind the kebab on desktop, with feature
parity in the existing mobile overflow menu
- Messages column anchors to max-w-3xl mx-auto to match the composer
- Chat bubbles drop from rounded-2xl to rounded-xl for vocabulary alignment
Typeset:
- Unify text sizing from 14 distinct sizes (with sub-pixel oddities and
rem/px duplicates) to a 5-step scale: 10px / 11px / text-xs / 13px / text-sm
WhatWeKnow collapsible:
- Header is now a toggle; section body hides when collapsed
- Auto-collapses on first render when facts ≥ 5 so Questions / Diagnostic
Checks stay above the fold
- Engineer's choice persists in sessionStorage per session and beats the
auto-collapse heuristic on subsequent renders
- key=activeChatId on both render sites resets state cleanly across sessions
Polish:
- Split MessageCircleQuestion into Pencil (question Answer CTA, write
affordance) + HelpCircle (per-check Explain toggle, universal help icon) —
same icon for two different jobs was a discoverability bug
- Drop redundant text-xs from font-sans text-[0.625rem] / text-[0.6875rem]
double-class definitions; the more-specific size always wins
TaskLane keyboard flow:
- Enter submits and auto-advances to the next pending task; Shift+Enter
inserts a newline (consistent across question and action textareas — paste
events don't fire keydown, so paste-then-Enter still works as expected)
- Esc cancels (same as the Cancel button)
- After the last pending task is submitted, focus moves to the Send Responses
button so the engineer can fire the whole batch with one more keystroke
- Subtle hint row under each open input teaches the shortcut
Type-check, lint, and build all clean.
Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
Superseded by ProposalBanner (Phase 8). The import was already removed
from AssistantChatPage in the previous commit; this deletes the orphaned
file itself and strips the now-unused suggestedFixSlot prop from
TaskLane's interface and both call sites.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Each lane section (What we know, Questions, Diagnostic Checks, Suggested
fix) had its own `position: sticky; top: 0` header. As the engineer
scrolled past a section, that section's header would pin until the
section's bottom edge cleared the viewport, producing an "orphaned"
label floating over unrelated content below. Headers now scroll with
their content — in a 340px-wide lane the affordance was negative value.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
- WhatWeKnow shows a "synthesizing" indicator + skeleton pulse while the
chat cycle is in-flight; task-lane header mirrors the signal with a
"thinking" pip so engineers know the AI is still working.
- Quiet-state hint when the lane is open (facts exist) but no open
questions, checks, or active fix — keeps the surface from looking
"finished" when the AI is about to follow up.
- Keyboard shortcuts: ⌘↵/Ctrl+↵ send in the composer (plain Enter still
sends), ⌘G toggles the Script Generator panel for the active fix,
`?` opens a new ShortcutsHelpOverlay listing all bindings. ⌘K palette
was already wired in TopBar.
- Responsive: below 1200px the task lane collapses to a bottom drawer
with a backdrop + a floating "Tasks ●" toggle button. TaskLane now
takes a `variant: 'side' | 'drawer'` prop; drawer variant drops the
resize handle and uses the shared slide-in-bottom animation.
- Build hygiene: fixed a pre-existing TS error in confirm-post error
handling (duplicate `response` type keys) and an unused-import warning
in TemplatizePrompt.
Verified: `npx tsc -b` and `npm run build` both clean against the dev
stack; Vite HMR applied each change without errors.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Adds the AI-proposed resolution path and the inline preview of the
markdown that will be posted to the customer ticket on Resolve. The
preview is keyed on (session_id, ai_sessions.state_version) so back-to-
back fetches against unchanged state hit an in-process cache instead
of paying for a Sonnet call.
Backend:
- preview_cache: in-process LRU keyed on (kind, session_id, state_version).
No TTL — state_version is the source of truth. Soft-cap 5000 entries.
- unified_chat_service: [SUGGEST_FIX] parser (last-block-wins, JSON
payload, confidence clamped 0-100), supersession persistence (sets
superseded_at on prior active row), atomic state_version bump.
- ResolutionNoteGeneratorService: pulls session, facts, active fix, and
redacted script_generations into a structured input bundle for Sonnet;
produces the four-section markdown (Problem / What we confirmed /
Root cause / Resolution). Sensitive script parameters redacted via
ScriptTemplateEngine.redact_sensitive driven by the template's
parameters_schema.
- /api/v1/ai-sessions/{id}/suggested-fixes/active — 200 with the active
fix or 404.
- /api/v1/ai-sessions/{id}/suggested-fixes/{fix_id}/decision — records
one_off / draft_template / build_template / dismissed; dismiss
supersedes; bumps state_version. 409 on dismissing an already-
superseded fix.
- /api/v1/ai-sessions/{id}/resolution-note/preview — generates or returns
cached markdown; from_cache flag in payload signals cache hit.
- scripts.py POST /generate now bumps state_version on the linked
ai_session_id when present (third source of preview-cache invalidation
per Section 5.5).
- ASSISTANT_SYSTEM_PROMPT documents [SUGGEST_FIX] (when to/not to emit,
format, supersession semantics).
- 12 tests covering the parser (well-formed, last-wins, malformed,
confidence clamping), supersession + state_version invariant, all
decision branches, preview cache hit-on-no-change + miss-after-write.
Frontend:
- src/components/pilot/sections/SuggestedFix.tsx — amber-accented card
with confidence badge; dismiss action wired to the decision endpoint.
- src/components/pilot/ResolutionNotePreview.tsx — popover with refresh,
loading state, cached/fresh indicator, ticket-ref display.
- src/api/sessionSuggestedFixes.ts — typed client; getActive normalizes
404 to null so callers don't have to special-case.
- TaskLane gains suggestedFixSlot + bottomSlot props (rendered after
Diagnostic Checks; bottomSlot anchors the Resolve action).
- AssistantChatPage: refreshSessionDerived helper batches fact + fix
refresh; fact mutations and chat sends both schedule a 500ms-debounced
preview refresh per the Section 5.5 spec.
Verified end-to-end against the dev stack with a real Sonnet call:
- /active 404 → fact create → preview generates four-section markdown
grounded only in provided facts → second preview call hits cache
(from_cache=true, no LLM call) → fact write 2 → cache miss, regenerates.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Adds the load-bearing structural feature of the FlowPilot migration: a
"What we know" panel that holds confirmed facts for a session, fed by AI
[PROMOTE] markers and engineer-added notes. Facts feed the resolution
note preview (Phase 3) and survive across turns via stable UUIDs assigned
to pending_task_lane items.
Backend:
- FactSynthesisService: create/update/soft-delete facts with atomic
state_version bumps; LLM-backed synthesize_from_question/check on the
fact_synthesis (Haiku) action tier per Section 6.6.
- /api/v1/ai-sessions/{id}/facts CRUD + /facts/promote (proposed_text or
via synthesis). PATCH returns 403 for question/diagnostic_check facts
(edit the source item instead, Section 7.3).
- unified_chat_service: [PROMOTE] marker parser (JSON-block per Section
8.1 spec drift note), stable-UUID assignment for pending_task_lane
questions/actions preserved by exact text/label match across turns.
- ASSISTANT_SYSTEM_PROMPT: documents [PROMOTE] format, when to/not to
emit, hallucination guardrails, source_ref handling.
- 17 tests covering parser, stable IDs, service validation, CRUD,
editability rule, both promote modes, 422 null-synthesis path,
state_version invariant.
Frontend:
- src/components/pilot/sections/{WhatWeKnow,WhatWeKnowItem,AddNoteButton}
— green-gradient section above Questions, dashed-circle check, inline
edit/delete gated by the server's editable flag.
- TaskLane gains a whatWeKnowSlot prop (existing assistant/ folder kept
per the doc's "rename is opportunistic" guidance).
- AssistantChatPage fetches facts on selectChat and refetches after each
chat send (so [PROMOTE]-synthesized facts appear immediately); auto-
opens the lane when facts exist.
Verification: end-to-end smoke against the local docker stack confirms
all five endpoints (list/create/patch/delete/promote) plus the 403
editability rule. pytest suite verifies the same with mocked LLM. Live
[PROMOTE] flow remains untested until used in the UI — the marker shape
is covered by parser tests.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Backend: start_session, prepare_session, batch_launch_sessions all missing
account_id=current_user.account_id — Phase 1 NOT NULL constraint made these
500 in test suite (test_ratings.py fixture couldn't create sessions).
Frontend ESLint:
- TaskLane.tsx: suppress react-refresh/only-export-components for clearTaskState
- TeamSummary.tsx: init loading from isAccountOwner to avoid sync setState in effect
- ScriptBodyEditor.tsx: move lastValueRef.current assignment into useEffect
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
- .gitignore: keep both graphify-out/ entries and main's .gitnexus entry
- ScriptCodeBlock/ScriptPreviewModal: take main's border-border and text-accent-text
for filename labels; use neutral ghost style for Save button in ScriptCodeBlock;
use bg-accent (normalized from bg-primary) for Save button in ScriptPreviewModal
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
- Import and call clearTaskState before updating questions/actions in
handleSend and handleTaskSubmit so new AI tasks always replace stale
sessionStorage cache instead of being overridden by it
- Include pending (not yet completed) tasks in the AI message on partial
submit so the AI knows which tasks were left unanswered
- Fix stale closure in TaskLane saveTaskLane useEffect — use refs for
questions/actions so the debounced backend save always uses current values
- Add responses field to pending_task_lane TypeScript type, removing the
unsafe double-cast in selectChat
- Instruct the AI to re-surface incomplete tasks unless ≥75% confident
the information is no longer needed
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Change "Questions" and "Diagnostic Checks" headers from text-muted
(#4f5666, barely visible on sidebar) to text-muted-foreground (#848b9b).
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Two issues fixed:
1. TaskLane useEffect on [questions, actions] was resetting all tasks
to pending with empty values, wiping restored user answers. Now
checks sessionStorage for saved state before resetting.
2. selectChat was setting activeQuestions/activeActions before writing
responses to sessionStorage, causing a race where TaskLane mounted
with new props but empty sessionStorage. Now writes responses to
sessionStorage first so TaskLane can restore them on prop change.
The backend saveTaskLane debounce (2s) persists responses to the DB,
and selectChat restores them via pending_task_lane.responses. This
chain now survives browser close.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Add PUT /ai-sessions/{id}/task-lane endpoint that saves the full task
lane state (AI questions/actions + user's in-progress responses) to
the pending_task_lane JSONB column. TaskLane debounce-saves to the
backend every 2s after changes. On session load, user responses are
restored from the backend into sessionStorage so TaskLane picks them
up on mount. Users can now close the browser, come back later, and
find their task lane exactly where they left it.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
TaskLane now saves user's in-progress answers (typed text, checked
items) to sessionStorage keyed by session ID. On reload or session
switch, the full task lane state restores — including partial work.
- TaskLane: saves tasks array to sessionStorage on every change,
restores from sessionStorage on mount
- AssistantChatPage: saves task lane metadata (visibility, questions,
actions, chatId) to sessionStorage, restores on mount
- Closing the task lane clears its sessionStorage entry
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
TaskLane had `if (submitted) return null` which immediately hid the
component after submit, before the AI could respond. Now the parent
controls visibility: the lane stays visible during the AI call, then
either updates with new tasks or clears when the AI sends none.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
eslint-disable-next-line only covers the next line, not lines further
down. Moved the comments from before useEffect() to before the actual
setState call inside the effect body.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
- Add back_populates to SessionBranch.session ↔ AISession.branches
to fix SQLAlchemy InvalidRequestError that broke all backend tests
- Remove unused 'tree' variable in command-palette.spec.ts
- Suppress setState-in-effect lint rule in TaskLane (intentional sync)
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Remove unused Wrench import from TaskLane, unused BranchMap import
and handleBranchSwitch from AssistantChatPage.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
- Enable submit when at least 1 item is answered (not all required)
- Dynamic label: "Send 2 of 6 Responses" vs "Send All Responses"
- Done cards are clickable to reopen for editing
- Collapsible preview shows formatted message before sending
- Resizable via left-edge grip handle, width persists to localStorage
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
