Add model selection, dashboard chat input, and model warming

- Add GET /api/chat/ps and POST /api/chat/warm endpoints for hot model
  visibility and pre-loading
- Extend PATCH /api/chat/conversations/:id to accept model in addition
  to title
- Add ModelSelector component with hot/cold indicators from Ollama /api/ps
- Add DashboardChatInput component (model selector + note picker + textarea)
  replacing the simple "New Chat" button on the dashboard
- Add model selector dropdown to ChatView header, persisted per-conversation
- Warm default model on dashboard mount via fire-and-forget background task
- Configure Ollama with OLLAMA_MAX_LOADED_MODELS=2 and OLLAMA_KEEP_ALIVE=30m
- Always-visible edit buttons on NoteCard/TaskCard (remove hover-only behavior)

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

docker-compose.yml

@@ -32,6 +32,9 @@ services:
     image: ollama/ollama
     volumes:
       - ollama_models:/root/.ollama
+    environment:
+      OLLAMA_MAX_LOADED_MODELS: "2"
+      OLLAMA_KEEP_ALIVE: "30m"
     # To enable GPU support, uncomment the deploy section below
     # (requires nvidia-container-toolkit)
     # deploy:

frontend/src/components/DashboardChatInput.vue

@@ -0,0 +1,358 @@
+<script setup lang="ts">
+import { ref, watch, nextTick } from "vue";
+import { apiGet } from "@/api/client";
+import { useChatStore } from "@/stores/chat";
+import ModelSelector from "@/components/ModelSelector.vue";
+import type { Note } from "@/types/note";
+
+const emit = defineEmits<{
+  submit: [payload: { content: string; model: string; contextNoteId?: number }];
+}>();
+
+const store = useChatStore();
+
+const selectedModel = ref(store.defaultModel || "");
+const messageInput = ref("");
+const inputEl = ref<HTMLTextAreaElement | null>(null);
+
+// Note picker state
+const attachedNote = ref<{ id: number; title: string } | null>(null);
+const showNotePicker = ref(false);
+const noteSearchQuery = ref("");
+const noteSearchResults = ref<{ id: number; title: string }[]>([]);
+const noteSearchLoading = ref(false);
+let noteSearchTimer: ReturnType<typeof setTimeout> | null = null;
+
+// Sync default model when store updates
+watch(
+  () => store.defaultModel,
+  (v) => {
+    if (!selectedModel.value && v) selectedModel.value = v;
+  }
+);
+
+function onSubmit() {
+  const content = messageInput.value.trim();
+  if (!content) return;
+  emit("submit", {
+    content,
+    model: selectedModel.value,
+    contextNoteId: attachedNote.value?.id,
+  });
+  messageInput.value = "";
+  attachedNote.value = null;
+  resetTextareaHeight();
+}
+
+function onInputKeydown(e: KeyboardEvent) {
+  if (e.key === "Enter" && !e.shiftKey) {
+    e.preventDefault();
+    onSubmit();
+  }
+}
+
+function autoResize() {
+  const el = inputEl.value;
+  if (!el) return;
+  el.style.height = "auto";
+  el.style.height = Math.min(el.scrollHeight, 120) + "px";
+}
+
+function resetTextareaHeight() {
+  const el = inputEl.value;
+  if (!el) return;
+  el.style.height = "auto";
+}
+
+// Note picker
+function toggleNotePicker() {
+  showNotePicker.value = !showNotePicker.value;
+  if (showNotePicker.value) {
+    noteSearchQuery.value = "";
+    noteSearchResults.value = [];
+    nextTick(() => {
+      const el = document.querySelector(
+        ".dashboard-chat .note-picker-search"
+      ) as HTMLInputElement;
+      el?.focus();
+    });
+  }
+}
+
+function onNoteSearchInput() {
+  if (noteSearchTimer) clearTimeout(noteSearchTimer);
+  noteSearchTimer = setTimeout(async () => {
+    const q = noteSearchQuery.value.trim();
+    if (!q) {
+      noteSearchResults.value = [];
+      return;
+    }
+    noteSearchLoading.value = true;
+    try {
+      const data = await apiGet<{ notes: Note[] }>(
+        `/api/notes?q=${encodeURIComponent(q)}&all=true&limit=5`
+      );
+      noteSearchResults.value = data.notes.map((n) => ({
+        id: n.id,
+        title: n.title,
+      }));
+    } catch {
+      noteSearchResults.value = [];
+    } finally {
+      noteSearchLoading.value = false;
+    }
+  }, 250);
+}
+
+function selectNote(note: { id: number; title: string }) {
+  attachedNote.value = note;
+  showNotePicker.value = false;
+}
+
+function removeAttachedNote() {
+  attachedNote.value = null;
+}
+</script>
+
+<template>
+  <div class="dashboard-chat">
+    <!-- Attached note pill -->
+    <div v-if="attachedNote" class="attached-note">
+      <span class="attached-note-pill">
+        {{ attachedNote.title }}
+        <button class="attached-note-remove" @click="removeAttachedNote">
+          &times;
+        </button>
+      </span>
+    </div>
+
+    <div class="chat-input-bar">
+      <ModelSelector v-model="selectedModel" :disabled="!store.chatReady" />
+
+      <div class="note-picker-wrapper">
+        <button
+          class="btn-attach"
+          @click="toggleNotePicker"
+          :disabled="!store.chatReady"
+          title="Attach a note"
+        >
+          <svg
+            width="18"
+            height="18"
+            viewBox="0 0 24 24"
+            fill="none"
+            stroke="currentColor"
+            stroke-width="2"
+            stroke-linecap="round"
+            stroke-linejoin="round"
+          >
+            <path
+              d="M21.44 11.05l-9.19 9.19a6 6 0 01-8.49-8.49l9.19-9.19a4 4 0 015.66 5.66l-9.2 9.19a2 2 0 01-2.83-2.83l8.49-8.48"
+            />
+          </svg>
+        </button>
+
+        <div v-if="showNotePicker" class="note-picker-dropdown">
+          <input
+            class="note-picker-search"
+            v-model="noteSearchQuery"
+            @input="onNoteSearchInput"
+            placeholder="Search notes..."
+          />
+          <div class="note-picker-results">
+            <div
+              v-for="note in noteSearchResults"
+              :key="note.id"
+              class="note-picker-item"
+              @click="selectNote(note)"
+            >
+              {{ note.title || "Untitled" }}
+            </div>
+            <div v-if="noteSearchLoading" class="note-picker-empty">
+              Searching...
+            </div>
+            <div
+              v-else-if="noteSearchQuery && !noteSearchResults.length"
+              class="note-picker-empty"
+            >
+              No notes found
+            </div>
+          </div>
+        </div>
+      </div>
+
+      <textarea
+        ref="inputEl"
+        v-model="messageInput"
+        @keydown="onInputKeydown"
+        @input="autoResize"
+        :placeholder="
+          store.chatReady
+            ? 'Start a new chat... (Enter to send)'
+            : 'Chat unavailable'
+        "
+        :disabled="!store.chatReady"
+        rows="1"
+      ></textarea>
+
+      <button
+        class="btn-send"
+        @click="onSubmit"
+        :disabled="!messageInput.trim() || !store.chatReady"
+      >
+        &uarr;
+      </button>
+    </div>
+  </div>
+</template>
+
+<style scoped>
+.dashboard-chat {
+  margin-top: 0.75rem;
+}
+
+.attached-note {
+  padding: 0.25rem 0;
+}
+.attached-note-pill {
+  display: inline-flex;
+  align-items: center;
+  gap: 0.25rem;
+  background: var(--color-primary);
+  color: #fff;
+  border-radius: 12px;
+  padding: 0.2rem 0.5rem;
+  font-size: 0.8rem;
+}
+.attached-note-remove {
+  background: none;
+  border: none;
+  color: rgba(255, 255, 255, 0.7);
+  cursor: pointer;
+  font-size: 1rem;
+  line-height: 1;
+  padding: 0 0.15rem;
+}
+.attached-note-remove:hover {
+  color: #fff;
+}
+
+.chat-input-bar {
+  display: flex;
+  align-items: center;
+  gap: 0.5rem;
+  padding: 0.5rem 0.5rem 0.5rem 0.75rem;
+  background: var(--color-input-bar-bg);
+  border-radius: 20px;
+  box-shadow: 0 2px 12px var(--color-shadow);
+}
+
+.chat-input-bar textarea {
+  flex: 1;
+  resize: none;
+  padding: 0.4rem 0.5rem;
+  border: none;
+  border-radius: 12px;
+  font-family: inherit;
+  font-size: 0.95rem;
+  background: transparent;
+  color: var(--color-input-bar-text);
+  outline: none;
+  max-height: 120px;
+  overflow-y: auto;
+}
+.chat-input-bar textarea::placeholder {
+  color: var(--color-input-bar-placeholder);
+}
+.chat-input-bar textarea:disabled {
+  opacity: 0.5;
+}
+
+/* Note picker */
+.note-picker-wrapper {
+  position: relative;
+}
+.btn-attach {
+  background: none;
+  border: none;
+  cursor: pointer;
+  color: var(--color-input-bar-text);
+  opacity: 0.6;
+  padding: 0.25rem;
+  display: flex;
+  align-items: center;
+  justify-content: center;
+}
+.btn-attach:hover {
+  opacity: 1;
+}
+.btn-attach:disabled {
+  opacity: 0.3;
+  cursor: default;
+}
+.note-picker-dropdown {
+  position: absolute;
+  bottom: calc(100% + 8px);
+  left: 0;
+  width: 280px;
+  background: var(--color-bg-card);
+  border: 1px solid var(--color-border);
+  border-radius: var(--radius-md);
+  box-shadow: 0 4px 16px var(--color-shadow);
+  z-index: 10;
+  overflow: hidden;
+}
+.note-picker-search {
+  width: 100%;
+  padding: 0.5rem 0.75rem;
+  border: none;
+  border-bottom: 1px solid var(--color-border);
+  background: transparent;
+  color: var(--color-text);
+  font-size: 0.9rem;
+  outline: none;
+  font-family: inherit;
+  box-sizing: border-box;
+}
+.note-picker-results {
+  max-height: 200px;
+  overflow-y: auto;
+}
+.note-picker-item {
+  padding: 0.5rem 0.75rem;
+  cursor: pointer;
+  font-size: 0.9rem;
+  overflow: hidden;
+  text-overflow: ellipsis;
+  white-space: nowrap;
+}
+.note-picker-item:hover {
+  background: var(--color-bg-secondary);
+}
+.note-picker-empty {
+  padding: 0.5rem 0.75rem;
+  color: var(--color-text-muted);
+  font-size: 0.85rem;
+}
+
+.btn-send {
+  width: 34px;
+  min-width: 34px;
+  height: 34px;
+  padding: 0;
+  display: flex;
+  align-items: center;
+  justify-content: center;
+  background: var(--color-primary);
+  color: #fff;
+  border: none;
+  border-radius: 50%;
+  cursor: pointer;
+  font-size: 1.1rem;
+  flex-shrink: 0;
+}
+.btn-send:disabled {
+  opacity: 0.35;
+  cursor: default;
+}
+</style>

frontend/src/components/ModelSelector.vue

@@ -0,0 +1,79 @@
+<script setup lang="ts">
+import { onMounted, computed } from "vue";
+import { useChatStore } from "@/stores/chat";
+
+defineProps<{
+  modelValue: string;
+  disabled?: boolean;
+}>();
+
+const emit = defineEmits<{
+  "update:modelValue": [value: string];
+}>();
+
+const store = useChatStore();
+
+function normalize(name: string): string {
+  return name.replace(/:latest$/, "");
+}
+
+const hotNames = computed(() => {
+  return new Set(store.runningModels.map((m) => normalize(m.name)));
+});
+
+function isHot(modelName: string): boolean {
+  return hotNames.value.has(normalize(modelName));
+}
+
+function onChange(e: Event) {
+  const val = (e.target as HTMLSelectElement).value;
+  emit("update:modelValue", val);
+}
+
+onMounted(async () => {
+  if (!store.models.length) store.fetchModels();
+  store.fetchRunningModels();
+});
+</script>
+
+<template>
+  <select
+    class="model-selector"
+    :value="modelValue"
+    :disabled="disabled"
+    @change="onChange"
+  >
+    <option v-if="!modelValue" value="" disabled>Select model</option>
+    <option
+      v-for="m in store.models"
+      :key="m.name"
+      :value="m.name"
+    >
+      {{ isHot(m.name) ? "\u2B24 " : "\u25CB " }}{{ m.name }}
+    </option>
+  </select>
+</template>
+
+<style scoped>
+.model-selector {
+  padding: 0.3rem 0.5rem;
+  border: 1px solid var(--color-border);
+  border-radius: var(--radius-sm);
+  background: var(--color-bg-secondary);
+  color: var(--color-text);
+  font-size: 0.85rem;
+  font-family: inherit;
+  cursor: pointer;
+  max-width: 220px;
+  overflow: hidden;
+  text-overflow: ellipsis;
+}
+.model-selector:disabled {
+  opacity: 0.5;
+  cursor: default;
+}
+.model-selector:focus {
+  outline: 2px solid var(--color-primary);
+  outline-offset: 1px;
+}
+</style>

frontend/src/components/NoteCard.vue

@@ -68,28 +68,19 @@ function goEdit() {
 }
 .btn-edit {
   flex-shrink: 0;
-  padding: 0.2rem 0.5rem;
-  font-size: 0.75rem;
-  background: transparent;
-  color: var(--color-text-muted);
+  padding: 0.25rem 0.6rem;
+  font-size: 0.8rem;
+  background: var(--color-bg-card);
+  color: var(--color-text-secondary);
   border: 1px solid var(--color-border);
   border-radius: var(--radius-sm);
   cursor: pointer;
-  opacity: 0;
-  transition: opacity 0.15s, color 0.15s, border-color 0.15s;
-}
-.note-card:hover .btn-edit {
-  opacity: 1;
+  transition: color 0.15s, border-color 0.15s, background 0.15s;
 }
 .btn-edit:hover {
   color: var(--color-primary);
   border-color: var(--color-primary);
 }
-@media (hover: none) {
-  .btn-edit {
-    opacity: 1;
-  }
-}
 .note-preview {
   margin: 0 0 0.5rem;
   color: var(--color-text-secondary);

frontend/src/components/TaskCard.vue

@@ -93,28 +93,19 @@ function isOverdue(): boolean {
 }
 .btn-edit {
   flex-shrink: 0;
-  padding: 0.2rem 0.5rem;
-  font-size: 0.75rem;
-  background: transparent;
-  color: var(--color-text-muted);
+  padding: 0.25rem 0.6rem;
+  font-size: 0.8rem;
+  background: var(--color-bg-card);
+  color: var(--color-text-secondary);
   border: 1px solid var(--color-border);
   border-radius: var(--radius-sm);
   cursor: pointer;
-  opacity: 0;
-  transition: opacity 0.15s, color 0.15s, border-color 0.15s;
-}
-.task-card:hover .btn-edit {
-  opacity: 1;
+  transition: color 0.15s, border-color 0.15s, background 0.15s;
 }
 .btn-edit:hover {
   color: var(--color-primary);
   border-color: var(--color-primary);
 }
-@media (hover: none) {
-  .btn-edit {
-    opacity: 1;
-  }
-}
 .task-preview {
   margin: 0 0 0.5rem;
   color: var(--color-text-secondary);

frontend/src/stores/chat.ts

@@ -15,6 +15,7 @@ import type {
   Message,
   OllamaModel,
   OllamaStatus,
+  RunningModel,
   SendMessageResponse,
 } from "@/types/chat";
 
@@ -27,6 +28,7 @@ export const useChatStore = defineStore("chat", () => {
   const streamingContent = ref("");
   const lastContextMeta = ref<ContextMeta | null>(null);
   const models = ref<OllamaModel[]>([]);
+  const runningModels = ref<RunningModel[]>([]);
   const ollamaStatus = ref<"checking" | "available" | "unavailable">("checking");
   const modelStatus = ref<"checking" | "ready" | "not_found">("checking");
   const defaultModel = ref("");
@@ -308,6 +310,42 @@ export const useChatStore = defineStore("chat", () => {
     }
   }
 
+  async function fetchRunningModels() {
+    try {
+      const data = await apiGet<{ models: RunningModel[] }>("/api/chat/ps");
+      runningModels.value = data.models;
+    } catch {
+      runningModels.value = [];
+    }
+  }
+
+  async function warmModel(model: string) {
+    try {
+      await apiPost("/api/chat/warm", { model });
+    } catch {
+      // Warming is best-effort
+    }
+  }
+
+  async function updateConversationModel(id: number, model: string) {
+    try {
+      const updated = await apiPatch<Conversation>(
+        `/api/chat/conversations/${id}`,
+        { model }
+      );
+      const idx = conversations.value.findIndex((c) => c.id === id);
+      if (idx !== -1) {
+        conversations.value[idx] = { ...conversations.value[idx], ...updated };
+      }
+      if (currentConversation.value?.id === id) {
+        currentConversation.value.model = updated.model;
+      }
+    } catch (e) {
+      useToastStore().show("Failed to update model", "error");
+      throw e;
+    }
+  }
+
   return {
     conversations,
     currentConversation,
@@ -317,6 +355,7 @@ export const useChatStore = defineStore("chat", () => {
     streamingContent,
     lastContextMeta,
     models,
+    runningModels,
     ollamaStatus,
     modelStatus,
     defaultModel,
@@ -331,6 +370,9 @@ export const useChatStore = defineStore("chat", () => {
     saveMessageAsNote,
     summarizeAsNote,
     fetchModels,
+    fetchRunningModels,
+    warmModel,
+    updateConversationModel,
     fetchStatus,
     startStatusPolling,
     stopStatusPolling,

frontend/src/types/chat.ts

@@ -31,6 +31,13 @@ export interface OllamaModel {
   size: number;
 }
 
+export interface RunningModel {
+  name: string;
+  size: number;
+  size_vram: number;
+  expires_at: string;
+}
+
 export interface ContextMeta {
   context_note_id: number | null;
   context_note_title: string | null;

frontend/src/views/ChatView.vue

@@ -6,6 +6,7 @@ import { useSettingsStore } from "@/stores/settings";
 import { apiGet } from "@/api/client";
 import { renderMarkdown } from "@/utils/markdown";
 import ChatMessage from "@/components/ChatMessage.vue";
+import ModelSelector from "@/components/ModelSelector.vue";
 import type { Note } from "@/types/note";
 
 const route = useRoute();
@@ -28,6 +29,9 @@ const noteSearchResults = ref<{ id: number; title: string }[]>([]);
 const noteSearchLoading = ref(false);
 let noteSearchTimer: ReturnType<typeof setTimeout> | null = null;
 
+// Model selection
+const selectedModel = ref("");
+
 // Exclude tracking (session-scoped per conversation)
 const excludedNoteIds = ref<Set<number>>(new Set());
 
@@ -71,6 +75,12 @@ onMounted(async () => {
   await store.fetchConversations();
   if (convId.value) {
     await store.fetchConversation(convId.value);
+    if (store.currentConversation?.model) {
+      selectedModel.value = store.currentConversation.model;
+    }
+  }
+  if (!selectedModel.value) {
+    selectedModel.value = store.defaultModel;
   }
   nextTick(() => inputEl.value?.focus());
 });
@@ -101,6 +111,26 @@ watch(convId, async (newId) => {
   nextTick(() => inputEl.value?.focus());
 });
 
+// Sync selectedModel from loaded conversation
+watch(
+  () => store.currentConversation?.model,
+  (model) => {
+    if (model) selectedModel.value = model;
+  }
+);
+
+// Persist model changes to backend
+watch(selectedModel, (newModel, oldModel) => {
+  if (
+    newModel &&
+    oldModel &&
+    newModel !== oldModel &&
+    store.currentConversation
+  ) {
+    store.updateConversationModel(store.currentConversation.id, newModel);
+  }
+});
+
 watch(
   () => store.streamingContent,
   () => scrollToBottom()
@@ -124,7 +154,7 @@ async function selectConversation(id: number) {
 }
 
 async function newConversation() {
-  const conv = await store.createConversation();
+  const conv = await store.createConversation("", selectedModel.value);
   await store.fetchConversation(conv.id);
   router.push(`/chat/${conv.id}`);
 }
@@ -313,6 +343,10 @@ function excludeAutoNote(noteId: number) {
               <line x1="3" y1="18" x2="21" y2="18"/>
             </svg>
           </button>
+          <ModelSelector
+            v-model="selectedModel"
+            :disabled="store.streaming"
+          />
           <h2>{{ store.currentConversation.title || "New Chat" }}</h2>
           <button
             v-if="store.currentConversation.messages.length"
@@ -584,6 +618,8 @@ function excludeAutoNote(noteId: number) {
   overflow: hidden;
   text-overflow: ellipsis;
   white-space: nowrap;
+  flex: 1;
+  min-width: 0;
 }
 
 .btn-summarize {

frontend/src/views/HomeView.vue

@@ -7,6 +7,7 @@ import type { Task, TaskListResponse } from "@/types/task";
 import type { Conversation } from "@/types/chat";
 import NoteCard from "@/components/NoteCard.vue";
 import TaskCard from "@/components/TaskCard.vue";
+import DashboardChatInput from "@/components/DashboardChatInput.vue";
 import type { TaskStatus } from "@/types/task";
 import { useTasksStore } from "@/stores/tasks";
 import { useChatStore } from "@/stores/chat";
@@ -166,11 +167,24 @@ function onStatusToggle(id: number, status: TaskStatus) {
   });
 }
 
-async function newChat() {
-  const chatStore = useChatStore();
-  const conv = await chatStore.createConversation();
+const chatStore = useChatStore();
+
+// Warm default model after status fetch
+chatStore.fetchStatus().then(() => {
+  if (chatStore.defaultModel) {
+    chatStore.warmModel(chatStore.defaultModel);
+  }
+});
+
+async function onChatSubmit(payload: {
+  content: string;
+  model: string;
+  contextNoteId?: number;
+}) {
+  const conv = await chatStore.createConversation("", payload.model);
   await chatStore.fetchConversation(conv.id);
   router.push(`/chat/${conv.id}`);
+  await chatStore.sendMessage(payload.content, payload.contextNoteId);
 }
 </script>
 
@@ -275,7 +289,7 @@ async function newChat() {
         <div v-else class="empty-state">
           <p class="empty-text">No chats yet.</p>
         </div>
-        <button class="btn-cta btn-new-chat" @click="newChat">+ New Chat</button>
+        <DashboardChatInput @submit="onChatSubmit" />
       </section>
 
       <section class="section">
@@ -374,9 +388,6 @@ async function newChat() {
   color: var(--color-text-muted);
   white-space: nowrap;
 }
-.btn-new-chat {
-  margin-top: 0.75rem;
-}
 .empty-state {
   text-align: center;
   padding: 1.5rem 0;

src/fabledassistant/routes/chat.py

@@ -15,7 +15,7 @@ from fabledassistant.services.chat import (
     list_conversations,
     save_response_as_note,
     summarize_conversation_as_note,
-    update_conversation_title,
+    update_conversation,
 )
 from fabledassistant.services.generation_buffer import (
     GenerationState,
@@ -83,9 +83,10 @@ async def update_conversation_route(conv_id: int):
     uid = get_current_user_id()
     data = await request.get_json()
     title = data.get("title")
-    if title is None:
-        return jsonify({"error": "title is required"}), 400
-    conv = await update_conversation_title(uid, conv_id, title)
+    model = data.get("model")
+    if title is None and model is None:
+        return jsonify({"error": "title or model is required"}), 400
+    conv = await update_conversation(uid, conv_id, title=title, model=model)
     if conv is None:
         return jsonify({"error": "Conversation not found"}), 404
     return jsonify(conv.to_dict())
@@ -233,6 +234,54 @@ async def summarize_conversation_route(conv_id: int):
         return jsonify({"error": str(e)}), 400
 
 
+@chat_bp.route("/ps", methods=["GET"])
+@login_required
+async def running_models_route():
+    """Return currently loaded (hot) models from Ollama."""
+    try:
+        async with httpx.AsyncClient(timeout=5.0) as client:
+            resp = await client.get(f"{Config.OLLAMA_URL}/api/ps")
+            resp.raise_for_status()
+            data = resp.json()
+            models = [
+                {
+                    "name": m["name"],
+                    "size": m.get("size", 0),
+                    "size_vram": m.get("size_vram", 0),
+                    "expires_at": m.get("expires_at", ""),
+                }
+                for m in data.get("models", [])
+            ]
+            return jsonify({"models": models})
+    except Exception as e:
+        logger.debug("Failed to query Ollama /api/ps: %s", e)
+        return jsonify({"models": []})
+
+
+@chat_bp.route("/warm", methods=["POST"])
+@login_required
+async def warm_model_route():
+    """Pre-load a model into Ollama memory."""
+    data = await request.get_json(force=True, silent=True) or {}
+    model = data.get("model", "").strip()
+    if not model:
+        return jsonify({"error": "model is required"}), 400
+
+    async def _warm():
+        try:
+            async with httpx.AsyncClient(timeout=120.0) as client:
+                await client.post(
+                    f"{Config.OLLAMA_URL}/api/generate",
+                    json={"model": model, "prompt": "", "keep_alive": "30m"},
+                )
+            logger.info("Warmed model %s", model)
+        except Exception as e:
+            logger.warning("Failed to warm model %s: %s", model, e)
+
+    asyncio.create_task(_warm())
+    return jsonify({"status": "warming"}), 202
+
+
 @chat_bp.route("/status", methods=["GET"])
 @login_required
 async def chat_status_route():

src/fabledassistant/services/chat.py

@@ -103,8 +103,11 @@ async def delete_conversation(user_id: int, conversation_id: int) -> bool:
         return True
 
 
-async def update_conversation_title(
-    user_id: int, conversation_id: int, title: str
+async def update_conversation(
+    user_id: int,
+    conversation_id: int,
+    title: str | None = None,
+    model: str | None = None,
 ) -> Conversation | None:
     async with async_session() as session:
         result = await session.execute(
@@ -116,13 +119,22 @@ async def update_conversation_title(
         conv = result.scalars().first()
         if conv is None:
             return None
-        conv.title = title
+        if title is not None:
+            conv.title = title
+        if model is not None:
+            conv.model = model
         conv.updated_at = datetime.now(timezone.utc)
         await session.commit()
         await session.refresh(conv)
         return conv
 
 
+async def update_conversation_title(
+    user_id: int, conversation_id: int, title: str
+) -> Conversation | None:
+    return await update_conversation(user_id, conversation_id, title=title)
+
+
 async def add_message(
     conversation_id: int,
     role: str,

src/fabledassistant/services/llm.py

@@ -80,7 +80,7 @@ async def stream_chat(
     payload: dict = {"model": model, "messages": messages, "stream": True}
     if options:
         payload["options"] = options
-    async with httpx.AsyncClient(timeout=httpx.Timeout(300.0, connect=10.0)) as client:
+    async with httpx.AsyncClient(timeout=httpx.Timeout(1800.0, connect=10.0, read=300.0)) as client:
         async with client.stream(
             "POST",
             f"{Config.OLLAMA_URL}/api/chat",
@@ -100,7 +100,7 @@ async def stream_chat(
 
 async def generate_completion(messages: list[dict], model: str) -> str:
     """Non-streaming chat completion, returns full response text."""
-    async with httpx.AsyncClient(timeout=httpx.Timeout(300.0, connect=10.0)) as client:
+    async with httpx.AsyncClient(timeout=httpx.Timeout(1800.0, connect=10.0, read=300.0)) as client:
         resp = await client.post(
             f"{Config.OLLAMA_URL}/api/chat",
             json={"model": model, "messages": messages, "stream": False},

summary.md

@@ -12,7 +12,7 @@
 >    Include file-level details in the commit body when the change is non-trivial.
 
 ## Last Updated
-2026-02-13 — Phase 5.9: Invitation system, table of contents, actionable dashboard, settings improvements
+2026-02-14 — Phase 6.0: Model selection, dashboard chat input, model warming
 
 ## Project Overview
 Fabled Assistant is a self-hosted note-taking and task-tracking application with
@@ -289,13 +289,13 @@ fabledassistant/
     │   │   ├── auth.ts          # Auth state: user, isAuthenticated, isAdmin, login/register/logout/checkAuth
     │   │   ├── notes.ts         # CRUD + tag filter, resolveTitle, convertToTask, convertToNote, fetchBacklinks, fetchAllTags (with toast errors)
     │   │   ├── tasks.ts         # CRUD + status/priority filter, patchStatus (with toast errors)
-    │   │   ├── chat.ts          # Conversation CRUD, sendMessage (SSE streaming), status polling (with toast errors)
+    │   │   ├── chat.ts          # Conversation CRUD, sendMessage (SSE streaming), status polling, running models, model warming, updateConversationModel (with toast errors)
     │   │   ├── settings.ts      # App settings: assistantName, defaultModel, pullModel, deleteModel (with toast errors)
     │   │   └── toast.ts         # Toast notification state (success/error/warning), 4s auto-dismiss, dismiss(id)
     │   ├── types/
     │   │   ├── auth.ts          # User interface, AuthStatus interface
     │   │   ├── note.ts          # Note interface (with status, priority, due_date, is_task) + TaskStatus, TaskPriority types + NoteListResponse
-    │   │   ├── chat.ts          # Message, Conversation, ConversationDetail, ContextMeta, OllamaModel, OllamaStatus interfaces
+    │   │   ├── chat.ts          # Message, Conversation, ConversationDetail, ContextMeta, OllamaModel, RunningModel, OllamaStatus interfaces
     │   │   ├── settings.ts      # AppSettings interface, ModelInfo interface (name, description, size, bestFor, category)
     │   │   └── task.ts          # Task = re-export of Note; TaskListResponse
     │   ├── extensions/
@@ -313,8 +313,8 @@ fabledassistant/
     │   │   ├── RegisterView.vue # Register form with password confirmation; shows "closed" message when registration disabled
     │   │   ├── RegisterInviteView.vue # Invitation-based registration: validates token, creates account with pre-set email
     │   │   ├── UserManagementView.vue # Admin user management: registration toggle, invitations (send/revoke), user list with delete
-    │   │   ├── ChatView.vue     # Dedicated /chat page: responsive sidebar (overlay on mobile), bubble messages, note picker, context pills
-    │   │   ├── HomeView.vue     # Actionable dashboard: overdue, due today, due this week, high priority, in progress tasks; recent chats; recently edited notes
+    │   │   ├── ChatView.vue     # Dedicated /chat page: responsive sidebar (overlay on mobile), bubble messages, note picker, context pills, model selector in header
+    │   │   ├── HomeView.vue     # Actionable dashboard: overdue, due today, due this week, high priority, in progress tasks; recent chats with inline chat input; recently edited notes; model warming on mount
     │   │   ├── SettingsView.vue # Settings page: assistant name, model catalog, data export/restore (admin)
     │   │   ├── NotesListView.vue # Note list: search, sort, tag filter pills, pagination
     │   │   ├── NoteEditorView.vue # Create/edit: Tiptap editor, sticky toolbar, AI Assist panel (bottom 1/3), Ctrl+S, unsaved guard
@@ -326,6 +326,8 @@ fabledassistant/
     │   │   ├── LogsView.vue     # Admin log viewer: stats summary, category/search/date filters, paginated table with expandable detail rows
     │   │   ├── AppHeader.vue    # Nav bar: brand, nav links (incl. admin Logs), status indicator, theme toggle, user info + logout, hamburger menu (mobile)
     │   │   ├── ChatPanel.vue    # Slide-out chat panel (right side overlay, receives contextNoteId prop), bubble-style messages, floating dark input, note picker, context pills with promote/exclude
+    │   │   ├── ModelSelector.vue # Model dropdown (v-model pattern): fetches installed + running models, hot/cold indicators
+    │   │   ├── DashboardChatInput.vue # Inline chat bar: ModelSelector + note picker + textarea + send button; emits submit event
     │   │   ├── ChatMessage.vue  # Message bubble: markdown rendering, configurable assistant name label, "Save as Note" action on assistant messages
     │   │   ├── NoteCard.vue     # Card with rendered markdown preview (v-html), TagPill, tag-click emit, hover edit button
     │   │   ├── TaskCard.vue     # Card with rendered preview (body not description), StatusBadge (clickable), PriorityBadge, due date, tags, hover edit button
@@ -393,12 +395,14 @@ fabledassistant/
 | POST | `/api/chat/conversations` | Create conversation (body: `{title?, model?}`) |
 | GET | `/api/chat/conversations/:id` | Get conversation with all messages |
 | DELETE | `/api/chat/conversations/:id` | Delete conversation (cascades to messages) |
-| PATCH | `/api/chat/conversations/:id` | Update conversation title (body: `{title}`) |
+| PATCH | `/api/chat/conversations/:id` | Update conversation title or model (body: `{title?, model?}`) |
 | POST | `/api/chat/conversations/:id/messages` | Start generation: save user message, launch background task, return 202 (body: `{content, context_note_id?, exclude_note_ids?}`) |
 | GET | `/api/chat/conversations/:id/generation/stream` | SSE endpoint tailing generation buffer; supports `Last-Event-ID` reconnection; emits `context`, `chunk`, `done`, `error` events |
 | POST | `/api/chat/conversations/:id/generation/cancel` | Cancel active generation (sets cancel_event, saves partial content) |
 | POST | `/api/chat/messages/:id/save-as-note` | Save assistant message as a new note (LLM-generated title, tagged `chat`) |
 | POST | `/api/chat/conversations/:id/summarize` | Summarize conversation via LLM, save as note |
+| GET | `/api/chat/ps` | List currently loaded (hot) Ollama models (`{models: [{name, size, size_vram, expires_at}]}`) |
+| POST | `/api/chat/warm` | Pre-load a model into Ollama memory (body: `{model}`, returns 202) |
 | GET | `/api/chat/status` | Check Ollama availability and model readiness (`{ollama, model, default_model}`) |
 | GET | `/api/chat/models` | List available Ollama models |
 | POST | `/api/chat/models/pull` | Pull/download a model from Ollama via SSE streaming progress (body: `{model}`, response: SSE with `{status, completed, total}` events) |
@@ -489,236 +493,84 @@ When adding a new migration, follow these conventions:
 - Set `down_revision` to chain from the previous migration's `revision`
 - Data migrations (like 0003) should use `op.get_bind()` + `sa.text()` for queries
 
-## Phased Roadmap
+## Implemented Features
 
-### Phase 1 — Skeleton & Dev Environment ✓
-- [x] Initialize Python project (pyproject.toml, Quart app scaffold)
-- [x] Initialize Vue.js project (Vite-based, inside `frontend/`)
-- [x] Set up Dockerfile (multi-stage: build Vue, serve with Quart)
-- [x] Docker Compose stack with Ollama service
-- [x] Quart serves Vue static build + `/api/health` endpoint
-- [x] Database setup (PostgreSQL 16, asyncpg, SQLAlchemy 2.0, Alembic)
+### Notes & Tasks
+- Full CRUD with markdown bodies, Obsidian-style `#tag` extraction (skips code fences), hierarchical tag filtering
+- Unified data model: a task is a note with `status IS NOT NULL` — convert freely between note ↔ task
+- Task attributes: status (todo/in_progress/done), priority (none/low/medium/high), due_date
+- Obsidian-style `[[wikilinks]]` with auto-create on click, backlinks ("what links here")
+- Tiptap WYSIWYG editor with markdown round-trip, tag/wikilink autocomplete and decorations, sticky toolbar
+- AI Assist panel in editor: background LLM generation via SSE with section targeting, accept/reject
+- Table of contents sidebar on note/task viewers (auto-generated from headings, hidden ≤1200px)
+- Inline edit buttons on NoteCard/TaskCard (hover on desktop, always visible on touch)
+- Search (ILIKE), sort, tag filter pills, pagination on list views
 
-### Phase 2 — Notes CRUD + UX ✓
-- [x] Database model for notes (title, body, tags[], parent_id, timestamps)
-- [x] REST API: create, read, update, delete, list notes with pagination
-- [x] Vue views: note list, note editor (markdown), note viewer (rendered)
-- [x] Search (ILIKE on title/body) and tag filtering (hierarchical via unnest)
-- [x] Inline `#tag` extraction from body text (backend regex, skips code fences)
-- [x] Hierarchical tag filtering (`#project` matches `project` and `project/*`)
-- [x] Dark/light theming with CSS custom properties + toggle + localStorage
-- [x] App header with navigation and theme toggle
-- [x] Tag pills (clickable + dismissible) on cards, viewer, and list filter bar
-- [x] Pagination bar with prev/next and page numbers
-- [x] Sort controls (field + asc/desc)
-- [x] Toast notifications (success/error, 3s auto-dismiss)
-- [x] Ctrl+S save shortcut in editor
-- [x] Unsaved changes guard (route leave + beforeunload)
-- [x] DOMPurify sanitization on rendered markdown
-- [x] Inline tag linkification in rendered markdown (clickable `#tag` links)
+### Dashboard
+- Actionable home page with 5 task sections: Overdue, Due Today, Due This Week, High Priority, In Progress
+- Priority-aware sorting (priority desc → due date asc) within each section
+- Cascading deduplication — tasks appear only in their highest-priority section
+- `due_before`/`due_after` query params on `/api/tasks` for date-range filtering
+- Recent chats and recently edited notes sections
+- All task sections hidden when empty; marking done removes from all lists
 
-### Phase 3 — Tasks CRUD + Wikilinks ✓
-- [x] Task model with status (todo/in_progress/done) and priority (none/low/medium/high) enums
-- [x] Alembic migration for tasks table with PG enums and indexes
-- [x] REST API: full CRUD + PATCH status toggle + filter by status/priority/tags
-- [x] Vue views: task list (search, status/priority filters, sort, pagination), editor (Ctrl+S, dirty guard), viewer (rendered markdown)
-- [x] StatusBadge (clickable, cycles status), PriorityBadge, TaskCard components
-- [x] Obsidian-style wikilinks: `[[Title]]` and `[[Title|Display]]` in rendered markdown
-- [x] Wikilink click handling resolves notes by title via `/api/notes/by-title`
+### LLM Chat
+- Ollama integration via async HTTP (httpx), auto-pull default model on startup
+- Background generation with `GenerationBuffer` (in-memory SSE fan-out, `Last-Event-ID` reconnect, 60s cleanup)
+- Stop generation with partial content preservation
+- Note-aware context building: current note + keyword search for related notes + URL fetching
+- Context pills with promote/exclude controls; note picker (paperclip) in chat input
+- LLM-generated conversation titles (re-generated every 10th message)
+- Save assistant messages as notes (LLM-titled, chat-tagged); summarize conversations
+- Dedicated `/chat` page with responsive sidebar + slide-out chat panel from header
+- Model catalog with installed/available tabs, download progress, select, remove
+- Per-conversation model selection via ModelSelector dropdown in chat header (persisted via PATCH)
+- Dashboard inline chat input: model selector + note picker + textarea; creates conversation and navigates
+- Model warming: default model pre-loaded into Ollama on dashboard mount via fire-and-forget POST
+- Hot/cold model indicators: `/api/chat/ps` proxies Ollama `/api/ps`; ModelSelector shows filled/empty circles
+- Ollama configured with `OLLAMA_MAX_LOADED_MODELS=2` and `OLLAMA_KEEP_ALIVE=30m`
 
-### Phase 3.5 — Note-Task Integration + Bug Fixes ✓
-- [x] **Wikilink auto-create:** Clicking `[[New Page]]` creates the note if it doesn't exist
-- [x] **Backlinks system:** "What links here" section on note and task viewers
-- [x] **Rendered markdown previews:** NoteCard/TaskCard show rendered markdown (links/images stripped)
-- [x] **Tag autocomplete Tab cycling:** Tab cycles through suggestions, single match accepts immediately
-- [x] **HomeView error isolation:** Notes and tasks load independently (one failure doesn't break both)
-- [x] **SPA routing fix:** Replaced catch-all route with 404 error handler to prevent API interception
-- [x] **500 error handler:** JSON error responses for API routes, traceback logging
-- [x] **Idempotent migrations:** All migrations rewritten to raw SQL with IF NOT EXISTS guards
-- [x] **Auto-migration on startup:** Dockerfile runs `alembic upgrade head` before starting app
+### Authentication & User Management
+- Session cookie auth with bcrypt, first-user-is-admin, orphaned data claiming
+- Per-user data isolation across all resources
+- Registration auto-closes after first user; admin toggle; invitation-based registration
+- Email invitation system: admin sends invite → branded email → `/register-invite` token flow
+- Password reset: email-based, SHA256-hashed tokens, 1-hour expiry
+- Session cookie hardening: HttpOnly, SameSite=Lax, optional Secure flag
+- Admin user management: list, delete, invite, revoke
 
-### Phase 3.6 — Merge Tasks into Notes ✓
-- [x] **Unified note/task model:** Task is just a note with `status IS NOT NULL`
-- [x] **Migration 0004:** Added `status`, `priority`, `due_date` columns to notes table, migrated task data from companion notes and orphan tasks, dropped `tasks` table
-- [x] **Eliminated companion note system:** No more companion note creation, title sync, cascade deletes, or `_skip_cascade` flags
-- [x] **Standardized on `body`:** Tasks use `body` everywhere (not `description`); API accepts `description` as fallback
-- [x] **Convert to task:** Simple `update_note(id, status='todo', priority='none')` — same ID preserved
-- [x] **Convert to note:** New `POST /api/notes/:id/convert-to-note` endpoint clears task attributes
-- [x] **Tasks service as thin wrappers:** `services/tasks.py` delegates entirely to `services/notes.py`
-- [x] **Frontend unified types:** `Task` is a re-export alias for `Note`; `note.ts` defines `TaskStatus`, `TaskPriority`
-- [x] **Simplified views:** Removed companion note UI from TaskEditorView/TaskViewerView/NoteViewerView; added Convert to Note button on TaskViewerView
+### Notifications & Email
+- SMTP email service (aiosmtplib): STARTTLS (587) and implicit TLS (465)
+- Security alerts: login, failed login, logout, password change (fire-and-forget)
+- Task due date reminders: hourly background check, grouped per user, dedup via logs
+- Invitation and password reset emails with branded HTML templates
+- Per-user notification preferences (task reminders, security alerts)
+- Admin SMTP configuration via Settings UI with test email
 
-### Phase 4 — LLM Chat Integration ✓
-- [x] **Ollama client:** async HTTP via httpx — `ensure_model()` (auto-pull on startup), `stream_chat()` (streaming NDJSON), `generate_completion()` (non-streaming)
-- [x] **Context building:** System prompt with note-aware context — includes current note (via `context_note_id`), keyword-extracted related note search (top 3, 5 keywords, 2000-char previews), URL content fetching (regex detection, HTML stripping, truncated to 4000 chars), returns `(messages, context_meta)` tuple with auto-found note IDs/titles, accepts `exclude_note_ids` to skip notes during auto-search
-- [x] **Conversation persistence:** `conversations` + `messages` tables in PostgreSQL, full CRUD
-- [x] **SSE streaming endpoint:** `POST /api/chat/conversations/:id/messages` streams LLM response chunks as SSE events, saves complete response to DB on finish
-- [x] **Save as note:** Save any assistant message as a new note (LLM-generated title, tagged `chat`)
-- [x] **Summarize conversation:** Send full conversation to LLM with summarize prompt, save result as note
-- [x] **Model management:** `GET /api/chat/models` lists available Ollama models; auto-pull default model (`llama3.1`) on app startup
-- [x] **Dedicated chat page:** `/chat` with sidebar (conversation list, create/delete) + message thread + markdown rendering + streaming indicator
-- [x] **Slide-out chat panel:** Toggle from header, receives `contextNoteId` from current route (`/notes/:id` or `/tasks/:id`), auto-creates conversation on first message
-- [x] **Frontend SSE client:** `apiStreamPost()` uses fetch + ReadableStream to parse SSE data lines
-- [x] **Auto-title:** Conversation title generated by LLM (concise 3-8 words)
+### Logging & Observability
+- Single `app_logs` table: audit (security events), usage (API requests), error (unhandled exceptions)
+- Request logging middleware with timing (skips log endpoints to avoid recursion)
+- Admin log viewer: stats summary, category/search/date filters, paginated table, expandable JSON details
+- Configurable retention via `LOG_RETENTION_DAYS` (default 90), hourly cleanup
 
-### Phase 4.5 — Chat UX + Settings + Model Management ✓
-- [x] **Settings infrastructure:** Key-value `settings` table, `GET/PUT /api/settings`, Pinia store with defaults
-- [x] **Configurable assistant name:** Default "Fable", editable in settings, injected into LLM system prompt and chat message labels
-- [x] **Settings page:** `/settings` route with assistant name form + model catalog
-- [x] **Model catalog:** 18 models across 3 categories (General Purpose, Coding, Uncensored / Creative Writing) with descriptions, sizes, and best-for labels
-- [x] **Model management:** Download (pull with SSE progress streaming + live percentage in UI), select (set as default), and remove (delete from Ollama) with confirmation UI
-- [x] **Status indicator in nav bar:** Moved from per-component (ChatView/ChatPanel) to global AppHeader; green/yellow/red dot with label
-- [x] **Chat bubble layout:** User messages right-aligned (primary color bg), assistant messages left-aligned (card bg), speech bubble tails
-- [x] **Floating dark input bar:** `#1c1c1e` background, rounded corners, circular send button with arrow
-- [x] **Auto-focus input:** Chat input auto-focuses on mount, conversation switch, and after sending
-- [x] **HTML entity fix:** `linkifyTags` regex now uses `(?<!&)` lookbehind to avoid matching `#39` inside `&#39;` as a tag. Additional layers: decode entities before marked, DOMPurify sanitization, explicit `&#39;` → `'` replacement after sanitization
-- [x] **New chat navigation fix:** `fetchConversation()` called before `router.push()` so `currentConversation` is set immediately
-- [x] **Recent chats on home page:** 3 most recent conversations shown as clickable cards + "New Chat" button
+### Settings & Admin
+- Per-user key-value settings store (assistant name, default model, notification prefs)
+- Admin: backup/restore (full or per-user JSON), SMTP config, base URL, registration toggle, user management, log viewer
+- Configurable base URL for email links (admin setting, env var fallback)
 
-### Phase 4.6 — Improved Note Context in Chat ✓
-- [x] **Multi-word search:** `list_notes()` splits multi-word queries into per-term ILIKE filters with AND logic (each word matched independently, not adjacent)
-- [x] **Context metadata:** `build_context()` returns `(messages, context_meta)` tuple with auto-found note IDs/titles and attached note info
-- [x] **Fuller auto-context:** Auto-found note previews increased from 300 to 2000 chars; uses all 5 extracted keywords instead of 3
-- [x] **Context SSE event:** Backend emits `{context: ContextMeta}` SSE event before streaming LLM response chunks
-- [x] **Exclude note IDs:** Backend accepts `exclude_note_ids` in request body, skips those notes during auto-search
-- [x] **Context pills:** ChatView and ChatPanel show auto-found notes as pills above streaming response with note title as router-link
-- [x] **Promote/exclude controls:** "+" button on context pills promotes note to attached context for next message; "x" button excludes note from future auto-searches in the session
-- [x] **Note picker:** Paperclip button in chat input opens dropdown with debounced search, selecting a note attaches it as context (shown as pill above input)
-- [x] **ContextMeta type:** New TypeScript interface for context metadata (`context_note_id`, `context_note_title`, `auto_notes`)
-- [x] **Session-scoped excludes:** Excluded note IDs tracked per conversation session, cleared on conversation switch
+### UI & Theming
+- Dark/light theme with CSS custom properties, design tokens, `prefers-color-scheme` detection
+- Responsive design: breakpoints (480/768/1024), hamburger menu, mobile touch targets (44px), sidebar overlays
+- App shell: navbar always visible, 100dvh flex layout, all views fit within viewport
+- Toast notifications (success/error/warning, 4s auto-dismiss)
+- DOMPurify sanitization on all rendered markdown
 
-### Phase 4.7 — Styling Consistency Pass ✓
-- [x] **Design tokens:** Added `--radius-sm/md/lg/pill`, `--color-success/warning/overlay`, `--focus-ring` to theme.css
-- [x] **Header alignment fix:** Removed `max-width: 960px` from header nav — now full-width with consistent padding, matching modern app layout
-- [x] **Active nav link state:** Added `router-link-active` styling with primary color + card background
-- [x] **Hardcoded colors replaced:** Status indicator dots, settings saved message, remove/confirm-delete buttons now use CSS variables
-- [x] **Border-radius standardized:** All components migrated from mixed 3px/4px/6px/8px to `var(--radius-sm)` (6px) for controls, `var(--radius-md)` (8px) for cards/containers
-- [x] **Button padding standardized:** All primary/CTA buttons use `0.45rem 1rem`; small buttons use `0.3rem 0.75rem`
-- [x] **Settings max-width fixed:** Changed from 700px to 720px to match all other views
-- [x] **Focus-visible states:** Global `focus-visible` rule with `--focus-ring` shadow on inputs, textareas, selects, buttons
-- [x] **Modal overlays:** Replaced hardcoded `rgba(0,0,0,0.5)` with `var(--color-overlay)` across all components
-
-### Phase 4.8 — Backend Efficiency & Consistency Pass ✓
-- [x] **`get_all_tags()` SQL rewrite:** Replaced O(n) Python-side tag extraction with single `SELECT DISTINCT unnest(tags)` query
-- [x] **`list_conversations()` subquery count:** Replaced `selectinload(Conversation.messages)` with correlated subquery for `message_count`, avoiding loading all messages for list view
-- [x] **Removed `services/tasks.py`:** Eliminated 64-line pass-through service; `routes/tasks.py` imports directly from `services.notes`
-- [x] **Consolidated convert functions:** `convert_note_to_task()` and `convert_task_to_note()` now use single session (was 3 DB round trips)
-- [x] **Batch settings updates:** New `set_settings_batch()` does all upserts in single transaction (was one transaction per setting)
-- [x] **Shared `get_installed_models()`:** Extracted duplicate Ollama model-listing logic from 3 locations into `services/llm.py`
-- [x] **`search_notes_for_context()`:** New function does single OR-keyword query (was up to 5 sequential `list_notes()` calls each with count query)
-- [x] **Logging:** Added `logger = logging.getLogger(__name__)` to `services/notes.py`, `services/chat.py`, `services/settings.py`
-- [x] **Database indexes:** Added B-tree indexes on `notes.title` (for `get_note_by_title` + backlinks) and `conversations.updated_at` (for list ordering)
-- [x] **Safe `Conversation.to_dict()`:** Handles case where messages relationship is not loaded
-
-### Phase 5 — Polish & Production Hardening ✓
-- [x] **Multi-user authentication:** Session cookie auth with bcrypt, first-user-is-admin pattern, orphaned data claiming
-- [x] **Per-user data isolation:** All notes, conversations, settings filtered by `user_id`
-- [x] **Auth decorators:** `@login_required`, `@admin_required` on all API routes (except health + auth status)
-- [x] **Frontend auth flow:** Login/register views, router beforeEach guard, ApiError with auto-redirect on 401
-- [x] **Error handling & logging:** Request logging with timing, sanitized 500 errors, LOG_LEVEL config, toast errors on all store actions
-- [x] **Toast improvements:** Warning type, dismiss button, 4s auto-dismiss
-- [x] **Responsive design:** Breakpoints (480/768/1024), utility classes (.hide-mobile/.hide-desktop), 44px mobile touch targets
-- [x] **Responsive header:** Hamburger menu at ≤768px, vertical nav dropdown, close on route change
-- [x] **Responsive chat:** Sidebar overlay on mobile with toggle button, slides in/out
-- [x] **Backup/restore:** Export user data (any user), full backup (admin), restore from JSON (admin)
-- [x] **Docker Swarm production stack:** `docker-compose.prod.yml` with Docker secrets, internal network isolation, health checks, resource limits, restart policies
-- [x] **Docker secrets support:** `_read_secret()` helper in config.py for SECRET_KEY_FILE, DATABASE_URL_FILE
-
-### Phase 5.1 — Chat UX Improvements ✓
-- [x] **LLM-generated titles:** Conversation titles generated by LLM (3-8 words) on first exchange and re-generated every 10th message to reflect evolved topics; falls back to truncation on failure
-- [x] **Background generation architecture:** `GenerationBuffer` (in-memory event buffer with SSE fan-out, reconnect via `Last-Event-ID`, auto-cleanup after 60s) + `run_generation()` asyncio task (streams from Ollama, periodic DB flushes every 5s)
-- [x] **Stop generation:** `cancel_event` on `GenerationBuffer`; `POST .../generation/cancel` endpoint; red stop button replaces send button during streaming; partial content saved as complete
-- [x] **Date/time in sidebar:** Relative time for recent (`Just now`, `5m ago`, `3h ago` up to 10h), then date (`Jan 15` this year, `Jan 15, 2025` older)
-- [x] **Empty chat cleanup:** Navigating away from an empty conversation (message_count === 0) auto-deletes it
-- [x] **Save-as-note LLM titles:** `save_response_as_note` generates title via LLM (same pattern as chat titles); falls back to first-line extraction on failure
-- [x] **Chat tag on saved notes:** Both `save_response_as_note` and `summarize_conversation_as_note` tag created notes with `chat`
-
-### Phase 5.2 — Tiptap Inline-Preview Editor + Layout Improvements ✓
-- [x] **Tiptap WYSIWYG editor:** Replaced plain textarea with Tiptap (ProseMirror-based) inline-preview editor — headings, bold, italic, lists, code blocks render inline while editing
-- [x] **Markdown round-trip:** Load markdown → HTML (via `marked`) → Tiptap editing → markdown (via custom `serializeToMarkdown()`) on every change
-- [x] **Tag/wikilink decorations:** ProseMirror decoration plugins visually highlight `#tag` and `[[wikilink]]` as colored badges without custom node types
-- [x] **Autocomplete migration:** Moved from textarea-based `useAutocomplete` to `@tiptap/suggestion` framework with shared `SuggestionDropdown` component
-- [x] **Heading vs tag disambiguation:** Tag suggestion suppressed at start of paragraphs (where `#` is likely a heading prefix)
-- [x] **Markdown paste handling:** Pasted markdown text auto-detected and converted to formatted content
-- [x] **Toolbar refactor:** `MarkdownToolbar` uses Tiptap commands with active state highlighting instead of textarea text insertion
-- [x] **Sticky toolbar:** Toolbar, title, tabs pinned above scrolling editor content
-- [x] **App shell layout:** Navbar always visible — `App.vue` uses flex column at 100dvh with header + content areas; all views fit below header
-- [x] **AI Assist panel:** Fixed at bottom 1/3 of viewport; section selector and prompt fill the panel; controls hidden during streaming/review (only output + accept/reject shown)
-- [x] **Auto-syncing sections:** `useAssist` watches `body` ref to keep section list in sync on load, typing, and AI accept
-- [x] **Accept trailing newline:** Accepted AI suggestions always end with a newline for clean separation
-- [x] **Wider content area:** Max-width increased from 720px to 960px across all views
-
-### Phase 5.3 — Registration Control, User Management, Security Hardening ✓
-- [x] **Registration auto-closes:** After the first user registers (admin), registration is closed by default
-- [x] **Registration toggle:** Admin can open/close registration from `/admin/users`
-- [x] **Password confirmation:** Register form requires password confirmation with inline validation
-- [x] **Admin user management:** `/admin/users` view with user list and delete (admin only)
-- [x] **Session cookie hardening:** `HttpOnly`, `SameSite=Lax` always on; `Secure` via `SECURE_COOKIES` env var
-- [x] **Default SECRET_KEY warning:** Logs WARNING on startup if default key is in use
-- [x] **Password change:** `PUT /api/auth/password` endpoint + Settings UI section
-- [x] **Production deployment docs:** README documents reverse proxy, rate limiting, CSP headers, SECRET_KEY, registration behavior
-
-### Phase 5.4 — Application Logging System ✓
-- [x] **Single `app_logs` table:** Unified audit/usage/error logging with `category` field
-- [x] **Usage logging middleware:** `after_request` logs all `/api/*` requests with user, endpoint, method, status, duration (skips `/api/admin/logs` to avoid recursion)
-- [x] **Error logging:** `handle_500` captures error type, message, and traceback
-- [x] **Audit logging:** Security events logged in auth routes (register, login, login_failed, logout, password_change) and admin routes (backup, restore, user_delete, registration_toggle)
-- [x] **Denormalized username:** Preserves attribution after user deletion (`user_id` FK uses `ON DELETE SET NULL`)
-- [x] **Admin log viewer:** `/admin/logs` with stats summary, category/search/date filters, paginated table with expandable JSON detail rows
-- [x] **Log stats endpoint:** `GET /api/admin/logs/stats` returns category counts
-- [x] **Configurable retention:** `LOG_RETENTION_DAYS` env var (default 90), hourly asyncio cleanup task
-- [x] **Config:** `LOG_RETENTION_DAYS` added to `Config` class
-
-### Phase 5.5 — SMTP Email Notifications ✓
-- [x] **SMTP email service:** `aiosmtplib` for async email sending, supports STARTTLS (587) and implicit TLS (465)
-- [x] **SMTP config in DB:** Admin configures SMTP via Settings UI, stored in `settings` table; env var fallbacks for Docker Swarm bootstrap
-- [x] **Admin SMTP endpoints:** `GET/PUT /api/admin/smtp` for config management, `POST /api/admin/smtp/test` for sending test emails
-- [x] **Security alert notifications:** Fire-and-forget `asyncio.create_task()` on login, failed login, logout, password change
-- [x] **Task due date reminders:** Hourly background loop checks for due/overdue tasks, sends grouped email per user; dedup via `app_logs` check
-- [x] **Per-user notification preferences:** `notify_task_reminders` and `notify_security_alerts` settings (default enabled)
-- [x] **Settings UI:** Notifications section for all users (checkbox toggles), SMTP section for admin (2-column grid + test email)
-
-### Phase 5.6 — Assist Background-Task + Buffer Architecture ✓
-- [x] **Assist buffer helpers:** `create_assist_buffer()`, `get_assist_buffer()`, `remove_assist_buffer()` using `"assist:{user_id}"` string keys in shared `_buffers` registry
-- [x] **`run_assist_generation()`:** Lightweight background task — streams from Ollama into buffer, no DB persistence, no title generation, no cancellation
-- [x] **Two-step assist API:** `POST /api/notes/assist` returns 202 + launches background task; `GET /api/notes/assist/stream` SSE endpoint tails buffer with 15s keepalives and `Last-Event-ID` reconnection
-- [x] **409 conflict guard:** Rejects concurrent assist requests per user
-- [x] **Frontend two-step pattern:** `useAssist.ts` calls `apiPost()` then `apiSSEStream()` with named event mapping (`chunk`/`done`/`error`); tracks stream handle for cleanup
-- [x] **Fixes `NS_ERROR_NET_PARTIAL_TRANSFER`:** Keepalive pings prevent browser/Hypercorn from closing connection during gaps between LLM chunks
-
-### Phase 5.7 — Invitation System ✓
-- [x] **Invitation tokens table:** `invitation_tokens` (migration 0012) — SHA256-hashed tokens, 7-day expiry, one-time use
-- [x] **Admin invitation management:** `POST/GET/DELETE /api/admin/invitations` for create/list/revoke
-- [x] **Invitation email:** Branded HTML email with accept link, sent via fire-and-forget asyncio task
-- [x] **Invitation registration flow:** `GET /api/auth/invitation/:token` validates, `POST /api/auth/register-with-invite` creates account
-- [x] **Frontend:** `/register-invite` view with token validation, email pre-fill, username/password form
-- [x] **Admin UI:** Invite User section in UserManagementView with email input, pending invitations table, revoke action
-- [x] **Base URL admin setting:** Configurable via Settings UI; used in invitation and password reset email links (replaces `Config.BASE_URL` hardcoded fallback)
-
-### Phase 5.8 — Table of Contents + Markdown Improvements ✓
-- [x] **TableOfContents component:** Sticky sidebar on note/task viewer — parses markdown headings, slugified IDs, smooth-scroll click, hidden on screens ≤1200px
-- [x] **Heading IDs in rendered markdown:** Custom `marked` renderer adds `id` attributes to headings for TOC anchor links
-- [x] **First-line tag stripping:** `stripFirstLineTags()` removes leading `#tag` lines from markdown before rendering (prevents tags from rendering as headings)
-- [x] **Viewer layout:** NoteViewerView and TaskViewerView use flex layout with main content + TOC sidebar (max-width 1200px)
-- [x] **Model catalog refresh:** Updated to current models (llama3.2/3.3, gemma3, qwen3, phi4, deepseek-r1, qwen2.5-coder, dolphin3), added Reasoning category, removed discontinued models
-- [x] **Settings model tabs:** Split model section into "Installed" and "Available" tabs
-
-### Phase 5.9 — Actionable "Today" Dashboard ✓
-- [x] **Due date filtering API:** Added `due_before` (exclusive `<`) and `due_after` (inclusive `>=`) params to `list_notes()` and `/api/tasks` endpoint
-- [x] **Dashboard rewrite:** HomeView shows 5 task sections: Overdue (red accent), Due Today, Due This Week (next 7 days), High Priority (amber accent, catches important tasks not already shown by date), In Progress (deduplicated catch-all)
-- [x] **Priority-aware sorting:** All task sections sort by priority desc then due date asc — highest priority items surface first within each section
-- [x] **Parallel fetching:** 6 API calls via `Promise.allSettled` — notes, overdue, due-soon (split client-side into today/week), high priority, in-progress, chats
-- [x] **Cascading deduplication:** Each section builds on a shared `seen` set so tasks only appear in their highest-priority section
-- [x] **Client-side filtering:** Filters out `done` tasks from all date/priority sections
-- [x] **Task sections hidden when empty:** Only shown if there are matching tasks
-- [x] **Status toggle:** Marking a task `done` removes it from all dashboard lists
-- [x] **Inline edit buttons:** NoteCard and TaskCard show an "Edit" button on hover (always visible on touch devices) that navigates directly to the edit view
-
-### Future / Stretch
-- Tagging/labeling system with LLM-suggested tags
-- Calendar/timeline view for tasks
-- Import/export (Markdown files, JSON)
-- Plugin/extension system
+### Infrastructure
+- Multi-stage Dockerfile: Node build → Python runtime, auto-migration on startup
+- Docker Compose (dev) + Docker Swarm production stack with secrets, overlay networks, health checks
+- Quart serves Vue SPA from static + REST API under `/api/`; 404 handler for SPA routing
+- Alembic migrations: 12 migrations, all raw SQL with idempotency guards (IF NOT EXISTS, DO $$ BEGIN...EXCEPTION)
+- Config from env vars + Docker secrets file support (`_read_secret`)
 
 ## Development Workflow
 - All development and testing done via Docker: `docker compose up --build`
@@ -728,37 +580,10 @@ When adding a new migration, follow these conventions:
   Quart serves them
 - To reset database: `docker compose down -v && docker compose up --build`
 
-## Current Status
-**Phase:** Phase 5.9 complete. Invitation system, table of contents, actionable dashboard, settings improvements.
-- Full note-taking and task-tracking CRUD with markdown, wikilinks, backlinks, tags
-- **Tiptap WYSIWYG editor** with inline formatting preview, markdown round-trip, paste handling
-- **Tag/wikilink autocomplete** via `@tiptap/suggestion` with heading disambiguation
-- Unified note/task model (a task is a note with `status IS NOT NULL`)
-- **Multi-user authentication** with session cookies, bcrypt passwords, admin role
-- **Per-user data isolation** across notes, conversations, and settings
-- LLM chat via Ollama with background generation task + SSE streaming
-- **AI Assist panel** pinned to bottom 1/3 of editor viewport with auto-syncing sections; uses background-task + buffer architecture with keepalive SSE (same pattern as chat)
-- Note-aware context: auto-includes current note + searches related notes by keyword
-- Context pills with promote/exclude controls; note picker in chat input
-- Save assistant messages as notes (LLM-titled, chat-tagged); summarize conversations as notes
-- Dedicated `/chat` page with responsive sidebar (overlay on mobile)
-- Settings page: assistant name, model catalog (installed/available tabs), password change, data export/restore (admin)
-- **Invitation system**: admin sends email invitations, recipients register via token-based `/register-invite` flow
-- **Registration control**: auto-closes after first user, admin toggle, invitation-based registration
-- **Admin user management**: `/admin/users` with user list, delete, invite, revoke
-- **Session cookie hardening**: HttpOnly, SameSite=Lax, optional Secure flag
-- **Actionable dashboard**: HomeView shows overdue/due-today/due-this-week/high-priority/in-progress task sections (priority-sorted, cascading dedup, hidden when empty), recent chats, recently edited notes
-- **Inline edit buttons**: NoteCard and TaskCard show hover edit button for direct navigation to edit view
-- **Table of contents**: Sticky sidebar on note/task viewers, auto-generated from markdown headings
-- **App-wide layout fix**: navbar always visible, all views fit within viewport
-- **Responsive design**: hamburger menu, mobile touch targets, responsive breakpoints
-- **Docker Swarm production stack** with secrets, network isolation, health checks, resource limits
-- **Application logging**: audit (security events), usage (API requests), error (unhandled exceptions)
-- **Admin log viewer**: `/admin/logs` with stats, filters, pagination, expandable detail rows
-- **Automatic log retention**: configurable via `LOG_RETENTION_DAYS` (default 90 days)
-- **SMTP email notifications**: security alerts (login/logout/failed login/password change), task due date reminders, invitation emails
-- **Admin SMTP configuration**: Settings UI with test email, DB-stored config with env var fallbacks
-- **Admin base URL setting**: Configurable public URL used in email links
-- **Per-user notification preferences**: toggle task reminders and security alerts independently
-- **Password reset flow**: Email-based with SHA256-hashed tokens, 1-hour expiry
-- Dark/light theme with CSS custom properties and design tokens
+## Backlog
+- Tagging/labeling system with LLM-suggested tags
+- Calendar/timeline view for tasks
+- Import/export (Markdown files, JSON)
+- Application-level rate limiting on auth endpoints
+- Security headers middleware (CSP, X-Frame-Options, etc.) — currently handled at reverse proxy level
+- Session invalidation on user deletion