feat(library): focused music/TV/movie landing screens + self-draining download queue

Library screens:
- Add dedicated music, TV, and movie landing pages (hero banner +
  horizontal carousels) backed by new music/tv/movies stores.
- Route tvshows libraries to /library/tv; surface rediscover ("haven't
  listened to in a while") albums via a new repository method across
  online/offline/hybrid repos plus the repository_get_rediscover_albums
  command.
- Add an A-Z jump bar for long alphabetically-sorted lists, with grid
  index anchors in LibraryGrid/LibraryListView/TrackList.
- Filter the "Podcasts" folder out of music library queries.

Downloads:
- Add a backend queue pump: enqueue_download / enqueue_video_downloads
  persist the resolved stream URL + target dir on each row (migration
  017), and the pump starts up to max_concurrent and drains the rest
  automatically as slots free, instead of the frontend silently dropping
  items past the concurrency limit. Album/series/season buttons now
  enqueue rather than calling start_download directly.

Other fixes:
- Hybrid search now returns instant cache results and pushes the merged
  cache+server union via a request-id-tagged search-event, so superseded
  queries can't clobber fresher results.
- URL-encode SearchTerm / genres / item types in online repo requests.
- Android: pause on audio-becoming-noisy (headphone/BT disconnect).

src/lib/api/bindings.ts

@@ -726,6 +726,31 @@ async markDownloadFailed(downloadId: number, errorMessage: string) : Promise<nul
 async startDownload(downloadId: number, streamUrl: string, targetDir: string) : Promise<null> {
     return await TAURI_INVOKE("start_download", { downloadId, streamUrl, targetDir });
 },
+/**
+ * Enqueue a download with its resolved stream URL, then let the queue pump
+ * start it (or a higher-priority pending item) when a slot is free.
+ * 
+ * Unlike [`start_download`], this never errors when the concurrency limit is
+ * reached: the URL is persisted on the row and the pump will pick it up once a
+ * slot frees. This is the path bulk operations (album/series/season) use so
+ * every queued item eventually downloads without the frontend re-issuing it.
+ */
+async enqueueDownload(downloadId: number, streamUrl: string, targetDir: string) : Promise<null> {
+    return await TAURI_INVOKE("enqueue_download", { downloadId, streamUrl, targetDir });
+},
+/**
+ * Enqueue a batch of already-queued video downloads, resolving each one's
+ * transcode URL from the repository using the `quality_preset` stored on the
+ * row. Then let the pump start them subject to the concurrency limit.
+ * 
+ * This is the bulk video path (series/season): `download_series`/
+ * `download_season` insert the rows, then this resolves URLs and enqueues them
+ * so they actually start. Resolving server-side avoids round-tripping every
+ * episode URL through the frontend.
+ */
+async enqueueVideoDownloads(handle: string, downloadIds: number[], targetDir: string) : Promise<null> {
+    return await TAURI_INVOKE("enqueue_video_downloads", { handle, downloadIds, targetDir });
+},
 /**
  * Get download manager statistics
  */
@@ -1037,6 +1062,12 @@ async repositoryGetRecentlyPlayedAudio(handle: string, limit: number | null) : P
 async repositoryGetResumeMovies(handle: string, limit: number | null) : Promise<MediaItem[]> {
     return await TAURI_INVOKE("repository_get_resume_movies", { handle, limit });
 },
+/**
+ * Get albums the user hasn't listened to recently ("rediscover")
+ */
+async repositoryGetRediscoverAlbums(handle: string, parentId: string | null, limit: number | null) : Promise<MediaItem[]> {
+    return await TAURI_INVOKE("repository_get_rediscover_albums", { handle, parentId, limit });
+},
 /**
  * Get genres for a library
  */
@@ -1046,8 +1077,8 @@ async repositoryGetGenres(handle: string, parentId: string | null) : Promise<Gen
 /**
  * Search for items
  */
-async repositorySearch(handle: string, query: string, options: SearchOptions | null) : Promise<SearchResult> {
-    return await TAURI_INVOKE("repository_search", { handle, query, options });
+async repositorySearch(handle: string, query: string, options: SearchOptions | null, requestId: number) : Promise<SearchResult> {
+    return await TAURI_INVOKE("repository_search", { handle, query, options, requestId });
 },
 /**
  * Get playback info for an item

src/lib/api/repository-client.test.ts

@@ -334,6 +334,7 @@ describe("RepositoryClient", () => {
           includeItemTypes: ["Audio"],
           limit: 100,
         },
+        requestId: 0,
       });
     });
   });

src/lib/api/repository-client.ts

@@ -109,12 +109,17 @@ export class RepositoryClient {
     return commands.repositoryGetResumeMovies(this.ensureHandle(), limit ?? null);
   }
 
+  /** Albums the user has played but not listened to recently ("rediscover"). */
+  async getRediscoverAlbums(parentId?: string, limit?: number): Promise<MediaItem[]> {
+    return commands.repositoryGetRediscoverAlbums(this.ensureHandle(), parentId ?? null, limit ?? null);
+  }
+
   async getGenres(parentId?: string): Promise<Genre[]> {
     return commands.repositoryGetGenres(this.ensureHandle(), parentId ?? null);
   }
 
-  async search(query: string, options?: SearchOptions): Promise<SearchResult> {
-    return commands.repositorySearch(this.ensureHandle(), query, options ?? null);
+  async search(query: string, options?: SearchOptions, requestId = 0): Promise<SearchResult> {
+    return commands.repositorySearch(this.ensureHandle(), query, options ?? null, requestId);
   }
 
   // ===== Playback Methods (all via Rust) =====

src/lib/components/library/AlbumDownloadButton.svelte

@@ -91,15 +91,18 @@
         // Get target directory for downloads
         const targetDir = await commands.storageGetPath();
 
-        // Start each queued track download
+        // Enqueue each track with its resolved stream URL. The backend queue
+        // pump starts up to max_concurrent at a time and advances through the
+        // rest automatically as slots free up — so we never hit (and silently
+        // drop) the concurrency limit the way startDownload did.
         for (let i = 0; i < tracks.length && i < downloadIds.length; i++) {
           try {
             const streamUrl = await repo.getAudioStreamUrl(tracks[i].id);
             if (streamUrl) {
-              await commands.startDownload(downloadIds[i], streamUrl, targetDir);
+              await commands.enqueueDownload(downloadIds[i], streamUrl, targetDir);
             }
           } catch (e) {
-            console.error(`Failed to start download for track ${tracks[i].id}:`, e);
+            console.error(`Failed to enqueue download for track ${tracks[i].id}:`, e);
           }
         }
 

src/lib/components/library/AlphabetScrollBar.svelte

@@ -0,0 +1,123 @@
+<!-- TRACES: UR-007 | DR-007 -->
+<script lang="ts">
+  /**
+   * A vertical A-Z index strip for long, alphabetically-sorted lists.
+   * Tapping or dragging a letter jumps to the first item that starts with it.
+   *
+   * The parent owns the actual scrolling: it passes `availableLetters`
+   * (which letters have items) and an `onJump(letter)` callback.
+   */
+
+  const ALPHABET = "ABCDEFGHIJKLMNOPQRSTUVWXYZ".split("");
+  const HASH = "#"; // bucket for names starting with a digit/symbol
+
+  interface Props {
+    /** Uppercase letters (or "#") that currently have at least one item. */
+    availableLetters: Set<string>;
+    /** Called with the chosen letter when the user picks one. */
+    onJump: (letter: string) => void;
+    /**
+     * CSS length reserved at the bottom of the viewport for the bottom nav /
+     * mini-player bars. The strip stretches to fill the space between the top
+     * sticky offset and this gap, so it ends just above those bars.
+     */
+    bottomGap?: string;
+  }
+
+  let { availableLetters, onJump, bottomGap = "5rem" }: Props = $props();
+
+  // The strip stretches to fill the height of its scroll container (the page's
+  // <main>, which sits below the sticky header) from the sticky top offset down
+  // to the bottom gap reserved for the nav / mini-player bars. We measure the
+  // container's visible height live so it stays correct regardless of header
+  // size, platform, or window resizes.
+  let container = $state<HTMLDivElement | null>(null);
+  let viewportHeight = $state(0);
+
+  function measure() {
+    const scroller = container?.closest("main") ?? document.documentElement;
+    viewportHeight = scroller.clientHeight;
+  }
+
+  $effect(() => {
+    measure();
+    const scroller = container?.closest("main");
+    const ro = new ResizeObserver(measure);
+    if (scroller) ro.observe(scroller);
+    window.addEventListener("resize", measure);
+    return () => {
+      ro.disconnect();
+      window.removeEventListener("resize", measure);
+    };
+  });
+
+  const letters = $derived([HASH, ...ALPHABET]);
+  let activeLetter = $state<string | null>(null);
+
+  function jumpTo(letter: string) {
+    if (!availableLetters.has(letter)) return;
+    activeLetter = letter;
+    onJump(letter);
+  }
+
+  // Allow dragging a finger/mouse down the strip to scrub through letters.
+  function letterAt(clientY: number): string | null {
+    const el = document.elementFromPoint(
+      // x is supplied by the caller via the bound container; we read from the
+      // element under the pointer instead to stay layout-agnostic.
+      lastClientX,
+      clientY
+    );
+    const letter = el?.getAttribute?.("data-letter");
+    return letter ?? null;
+  }
+
+  let lastClientX = $state(0);
+  let isScrubbing = $state(false);
+
+  function handlePointerDown(e: PointerEvent) {
+    isScrubbing = true;
+    lastClientX = e.clientX;
+    const letter = letterAt(e.clientY);
+    if (letter) jumpTo(letter);
+  }
+
+  function handlePointerMove(e: PointerEvent) {
+    if (!isScrubbing) return;
+    lastClientX = e.clientX;
+    const letter = letterAt(e.clientY);
+    if (letter && letter !== activeLetter) jumpTo(letter);
+  }
+
+  function handlePointerUp() {
+    isScrubbing = false;
+  }
+</script>
+
+<svelte:window onpointerup={handlePointerUp} onpointercancel={handlePointerUp} />
+
+<div
+  bind:this={container}
+  class="flex flex-col items-center justify-between select-none touch-none py-1"
+  style="height: calc({viewportHeight}px - 1.5rem - {bottomGap})"
+  onpointerdown={handlePointerDown}
+  onpointermove={handlePointerMove}
+  role="navigation"
+  aria-label="Jump to letter"
+>
+  {#each letters as letter (letter)}
+    {@const enabled = availableLetters.has(letter)}
+    <button
+      type="button"
+      data-letter={letter}
+      disabled={!enabled}
+      onclick={() => jumpTo(letter)}
+      class="w-5 leading-tight text-[10px] sm:text-xs font-semibold transition-colors
+        {enabled ? 'text-gray-400 hover:text-[var(--color-jellyfin)]' : 'text-gray-700 cursor-default'}
+        {activeLetter === letter && enabled ? 'text-[var(--color-jellyfin)] scale-125' : ''}"
+      aria-label={`Jump to ${letter}`}
+    >
+      {letter}
+    </button>
+  {/each}
+</div>

src/lib/components/library/GenericMediaListPage.svelte

@@ -4,6 +4,8 @@
   import { goto } from "$app/navigation";
   import { currentLibrary } from "$lib/stores/library";
   import { auth } from "$lib/stores/auth";
+  import { shouldShowAudioMiniPlayer } from "$lib/stores/player";
+  import { isAndroid } from "$lib/stores/appState";
   import SearchBar from "$lib/components/common/SearchBar.svelte";
   import SortButtonGroup from "$lib/components/common/SortButtonGroup.svelte";
   import type { SortOption } from "$lib/components/common/SortButtonGroup.svelte";
@@ -13,6 +15,8 @@
   import type { MediaItem, Library, ItemType } from "$lib/api/types";
   import LibraryGrid from "./LibraryGrid.svelte";
   import TrackList from "./TrackList.svelte";
+  import AlphabetScrollBar from "./AlphabetScrollBar.svelte";
+  import { excludePodcasts } from "$lib/utils/podcastFilter";
 
   /**
    * Generic media list page supporting Albums, Artists, Playlists, and Tracks
@@ -41,6 +45,7 @@
 
   let items = $state<MediaItem[]>([]);
   let loading = $state(true);
+  let gridWrapper = $state<HTMLDivElement | null>(null);
   let searchQuery = $state("");
   let debouncedSearchQuery = $state("");
   let sortBy = $state<string>("");
@@ -74,12 +79,13 @@
       const repo = auth.getRepository();
 
       // Use backend search if search query is provided, otherwise use getItems with sort
+      // HACK: excludePodcasts drops the "Podcasts" folder stored in the music library.
       if (debouncedSearchQuery.trim()) {
         const result = await repo.search(debouncedSearchQuery, {
           includeItemTypes: [config.itemType],
           limit: 10000,
         });
-        items = result.items;
+        items = excludePodcasts(result.items);
       } else {
         const result = await repo.getItems($currentLibrary.id, {
           includeItemTypes: [config.itemType],
@@ -88,7 +94,7 @@
           recursive: true,
           limit: 10000,
         });
-        items = result.items;
+        items = excludePodcasts(result.items);
       }
     } catch (e) {
       console.error(`Failed to load ${config.itemType}:`, e);
@@ -142,6 +148,54 @@
       goto(`/library/${track.id}`);
     }
   }
+
+  // ===== A-Z jump bar =====
+  // Bucket a name to its index letter: A-Z, or "#" for digits/symbols/empty.
+  function letterFor(name: string): string {
+    const first = (name ?? "").trim().charAt(0).toUpperCase();
+    return first >= "A" && first <= "Z" ? first : "#";
+  }
+
+  // Only meaningful when the list is sorted alphabetically and long enough to scroll.
+  const isAlphaSorted = $derived(sortBy === "SortName");
+  const showAlphaBar = $derived(
+    isAlphaSorted &&
+      !loading &&
+      !debouncedSearchQuery.trim() &&
+      items.length > 30
+  );
+
+  const availableLetters = $derived.by(() => {
+    const set = new Set<string>();
+    if (showAlphaBar) {
+      for (const item of items) set.add(letterFor(item.name));
+    }
+    return set;
+  });
+
+  // First item index for each letter, honouring current ascending/descending order.
+  const firstIndexForLetter = $derived.by(() => {
+    const map = new Map<string, number>();
+    items.forEach((item, index) => {
+      const letter = letterFor(item.name);
+      if (!map.has(letter)) map.set(letter, index);
+    });
+    return map;
+  });
+
+  function jumpToLetter(letter: string) {
+    const index = firstIndexForLetter.get(letter);
+    if (index === undefined || !gridWrapper) return;
+    const target = gridWrapper.querySelector(`[data-grid-index="${index}"]`);
+    target?.scrollIntoView({ behavior: "smooth", block: "start" });
+  }
+
+  // Bottom space the layout's <main> reserves for the nav / mini-player bars.
+  // Mirrors src/routes/library/+layout.svelte so the A-Z strip ends just above
+  // whichever bars are visible.
+  const bottomGap = $derived(
+    $shouldShowAudioMiniPlayer ? ($isAndroid ? "11rem" : "7rem") : "5rem"
+  );
 </script>
 
 <div class="space-y-6">
@@ -192,10 +246,19 @@
       <p>No {config.title.toLowerCase()} found</p>
     </div>
   {:else}
-    {#if config.displayComponent === "grid"}
-      <LibraryGrid items={items} onItemClick={handleItemClick} musicContent={["MusicAlbum", "MusicArtist", "Audio", "Playlist"].includes(config.itemType)} />
-    {:else if config.displayComponent === "tracklist"}
-      <TrackList tracks={items} onTrackClick={handleTrackClick} />
-    {/if}
+    <div class="flex gap-2">
+      <div bind:this={gridWrapper} class="flex-1 min-w-0">
+        {#if config.displayComponent === "grid"}
+          <LibraryGrid items={items} onItemClick={handleItemClick} musicContent={["MusicAlbum", "MusicArtist", "Audio", "Playlist"].includes(config.itemType)} />
+        {:else if config.displayComponent === "tracklist"}
+          <TrackList tracks={items} onTrackClick={handleTrackClick} />
+        {/if}
+      </div>
+      {#if showAlphaBar}
+        <div class="sticky top-2 self-start flex-shrink-0 h-fit">
+          <AlphabetScrollBar {availableLetters} onJump={jumpToLetter} {bottomGap} />
+        </div>
+      {/if}
+    </div>
   {/if}
 </div>

src/lib/components/library/LibraryGrid.svelte

@@ -69,12 +69,14 @@
     <LibraryListView {items} showProgress={true} onItemClick={onItemClick} />
   {:else}
     <div class="grid grid-cols-2 sm:grid-cols-3 md:grid-cols-4 lg:grid-cols-5 xl:grid-cols-6 gap-4">
-      {#each items as item (item.id)}
-        <MediaCard
-          {item}
-          showProgress={true}
-          onclick={() => onItemClick?.(item)}
-        />
+      {#each items as item, index (item.id)}
+        <div data-grid-index={index}>
+          <MediaCard
+            {item}
+            showProgress={true}
+            onclick={() => onItemClick?.(item)}
+          />
+        </div>
       {/each}
     </div>
   {/if}

src/lib/components/library/LibraryListView.svelte

@@ -68,6 +68,7 @@
 
     <button
       type="button"
+      data-grid-index={index}
       onclick={() => onItemClick?.(item)}
       class="w-full flex items-center gap-3 p-2 rounded-lg hover:bg-[var(--color-surface)] transition-colors group"
     >

src/lib/components/library/SeasonDownloadButton.svelte

@@ -71,6 +71,13 @@
 
       // Pin the season item
       await downloads.pinItem(seasonId);
+
+      // Resolve each episode's transcode URL (server-side) and enqueue. The
+      // backend pump then starts up to max_concurrent and advances through the
+      // rest as slots free up.
+      const handle = auth.getRepository().getHandle();
+      await commands.enqueueVideoDownloads(handle, downloadIds, targetDir);
+      console.log("   Episodes enqueued; backend pump will start them");
     } catch (error) {
       console.error("Failed to start season download:", error);
     } finally {

src/lib/components/library/SeriesDownloadButton.svelte

@@ -64,9 +64,12 @@
       // Pin the series item
       await downloads.pinItem(seriesId);
 
-      // Start downloads (the backend will handle queuing)
-      // For now, we'll rely on a download manager to pick them up
-      // TODO: Implement batch download start
+      // Resolve each episode's transcode URL (server-side) and enqueue. The
+      // backend pump then starts up to max_concurrent and advances through the
+      // rest as slots free up.
+      const handle = auth.getRepository().getHandle();
+      await commands.enqueueVideoDownloads(handle, downloadIds, targetDir);
+      console.log("   Episodes enqueued; backend pump will start them");
     } catch (error) {
       console.error("Failed to start series download:", error);
     } finally {

src/lib/components/library/TrackList.svelte

@@ -204,7 +204,7 @@
   <!-- Track Rows -->
   <div class="space-y-1">
     {#each tracks as track, index (track.id)}
-      <div class="w-full group hover:bg-[var(--color-surface-hover)] rounded-lg transition-colors relative {currentlyPlayingId === track.id ? 'bg-[var(--color-jellyfin)]/10 border-l-4 border-[var(--color-jellyfin)]' : ''}">
+      <div data-grid-index={index} class="w-full group hover:bg-[var(--color-surface-hover)] rounded-lg transition-colors relative {currentlyPlayingId === track.id ? 'bg-[var(--color-jellyfin)]/10 border-l-4 border-[var(--color-jellyfin)]' : ''}">
         <!-- Desktop View -->
         <button
           onclick={() => handleTrackClick(track, index)}

src/lib/stores/library.ts

@@ -2,9 +2,19 @@
 // TRACES: UR-007, UR-008, UR-029, UR-030 | DR-007, DR-011, DR-033
 
 import { writable, derived } from "svelte/store";
+import { listen, type UnlistenFn } from "@tauri-apps/api/event";
 import type { Library, MediaItem, SearchResult, Genre } from "$lib/api/types";
 import { auth } from "./auth";
 
+/**
+ * Payload of the backend `search-event` (mirrors Rust `SearchUpdateEvent`).
+ * Carries the merged cache+server results for a given search request.
+ */
+interface SearchUpdateEvent {
+  requestId: number;
+  result: SearchResult;
+}
+
 export type ViewMode = "grid" | "list";
 
 interface LibraryState {
@@ -47,8 +57,23 @@ function createLibraryStore() {
 
   const { subscribe, set, update } = writable<LibraryState>(initialState);
 
-  // Test log to confirm cache logging is active
-  console.log("✅ [LibraryStore] Cache logging enabled - you should see cache hit/miss logs below");
+  // Monotonic id identifying the most recent search request. Each new search
+  // bumps it; the deferred `search-event` (carrying merged cache+server
+  // results) is only applied when its requestId still matches the latest one,
+  // so out-of-order / superseded results never clobber fresher ones.
+  let searchRequestId = 0;
+  let unlistenSearch: UnlistenFn | null = null;
+
+  // Lazily subscribe to backend search updates the first time we search.
+  async function ensureSearchListener() {
+    if (unlistenSearch) return;
+    unlistenSearch = await listen<SearchUpdateEvent>("search-event", (event) => {
+      const { requestId, result } = event.payload;
+      // Ignore results from a query the user has already moved on from.
+      if (requestId !== searchRequestId) return;
+      update((s) => ({ ...s, searchResults: result.items }));
+    });
+  }
 
   async function loadLibraries() {
     update((s) => ({ ...s, loadingCount: s.loadingCount + 1, error: null }));
@@ -157,11 +182,17 @@ function createLibraryStore() {
   }
 
   async function search(query: string) {
+    // Bump the request id for every call (including clears) so any in-flight
+    // backend update for a previous query is ignored when it arrives.
+    const requestId = ++searchRequestId;
+
     if (!query.trim()) {
       update((s) => ({ ...s, searchQuery: "", searchResults: [] }));
       return;
     }
 
+    await ensureSearchListener();
+
     update((s) => ({ ...s, loadingCount: s.loadingCount + 1, error: null, searchQuery: query }));
 
     try {
@@ -172,16 +203,21 @@ function createLibraryStore() {
         setTimeout(() => reject(new Error("Search timeout - please try again")), 10000)
       );
 
+      // Phase 1: the command resolves with instant local-cache results. The
+      // merged (cache + server) union arrives later via the `search-event`
+      // listener above, tagged with this same requestId.
       const result = await Promise.race([
-        repo.search(query, { limit: 10000 }),
+        repo.search(query, { limit: 10000 }, requestId),
         timeoutPromise
       ]);
 
-      update((s) => ({
-        ...s,
-        searchResults: result.items,
-        loadingCount: Math.max(0, s.loadingCount - 1),
-      }));
+      // Only apply if this is still the active query (a newer search may have
+      // started while we awaited).
+      if (requestId === searchRequestId) {
+        update((s) => ({ ...s, searchResults: result.items }));
+      }
+
+      update((s) => ({ ...s, loadingCount: Math.max(0, s.loadingCount - 1) }));
 
       return result;
     } catch (error) {
@@ -196,6 +232,8 @@ function createLibraryStore() {
   }
 
   function clearSearch() {
+    // Invalidate any in-flight backend search update.
+    searchRequestId++;
     update((s) => ({ ...s, searchQuery: "", searchResults: [] }));
   }
 
@@ -247,6 +285,11 @@ function createLibraryStore() {
   }
 
   function reset() {
+    searchRequestId++;
+    if (unlistenSearch) {
+      unlistenSearch();
+      unlistenSearch = null;
+    }
     set(initialState);
   }
 

src/lib/stores/movies.ts

@@ -0,0 +1,161 @@
+// Movies library landing page data store.
+// Powers the focused movies landing: hero + horizontal sliders.
+// TRACES: UR-007, UR-023, UR-034 | DR-007, DR-038, DR-039
+import { writable, derived } from "svelte/store";
+import type { MediaItem } from "$lib/api/types";
+import { auth } from "./auth";
+
+/** A single "by genre" row: the genre name plus the movies in it. */
+export interface GenreRow {
+  id: string;
+  name: string;
+  items: MediaItem[];
+}
+
+interface MoviesState {
+  // Movies the user can resume (continue watching).
+  continueWatching: MediaItem[];
+  // Recently added movies in the library.
+  recentlyAdded: MediaItem[];
+  // One slider per genre (top genres by movie count).
+  genreRows: GenreRow[];
+  // Mix used for the hero banner.
+  heroItems: MediaItem[];
+  isLoading: boolean;
+  error: string | null;
+}
+
+const SECTION_LIMIT = 16;
+// How many genre sliders to show, and how many genres to probe to find them.
+const MAX_GENRE_ROWS = 8;
+const MAX_GENRES_PROBED = 20;
+
+function createMoviesStore() {
+  const initialState: MoviesState = {
+    continueWatching: [],
+    recentlyAdded: [],
+    genreRows: [],
+    heroItems: [],
+    isLoading: false,
+    error: null,
+  };
+
+  const { subscribe, set, update } = writable<MoviesState>(initialState);
+
+  /**
+   * Build the hero rotation. Prefer in-progress movies (most personal), then
+   * fall back to recently added. De-duplicates by id and prefers items that
+   * carry backdrop/primary artwork for a good banner.
+   */
+  function buildHero(continueWatching: MediaItem[], recentlyAdded: MediaItem[]): MediaItem[] {
+    const hasArt = (i: MediaItem) =>
+      !!(i.backdropImageTags && i.backdropImageTags.length > 0) || !!i.primaryImageTag;
+
+    const result: MediaItem[] = [];
+    const seen = new Set<string>();
+
+    for (const pool of [continueWatching, recentlyAdded]) {
+      for (const item of pool) {
+        if (result.length >= 6) break;
+        if (hasArt(item) && !seen.has(item.id)) {
+          seen.add(item.id);
+          result.push(item);
+        }
+      }
+    }
+
+    return result.slice(0, 6);
+  }
+
+  async function loadSections(libraryId: string) {
+    update(s => ({
+      ...s,
+      isLoading: s.continueWatching.length === 0 && s.recentlyAdded.length === 0,
+      error: null,
+    }));
+
+    try {
+      const repo = auth.getRepository();
+
+      const [resume, latest] = await Promise.all([
+        repo.getResumeMovies(SECTION_LIMIT),
+        repo.getLatestItems(libraryId, SECTION_LIMIT),
+      ]);
+
+      const heroItems = buildHero(resume, latest);
+
+      update(s => ({
+        ...s,
+        continueWatching: resume,
+        recentlyAdded: latest,
+        heroItems,
+        isLoading: false,
+      }));
+
+      // Genre rows are secondary — load them after the main sections paint so
+      // the page isn't blocked on N per-genre queries.
+      loadGenreRows(libraryId);
+    } catch (error) {
+      const message = error instanceof Error ? error.message : "Failed to load movie sections";
+      update(s => ({ ...s, isLoading: false, error: message }));
+      console.error("Failed to load movie sections:", error);
+    }
+  }
+
+  /**
+   * Build one slider per genre, showing the top-rated movies in each. We probe
+   * a bounded set of genres in parallel, drop empty ones, then keep the genres
+   * with the most movies (so niche/near-empty genres don't crowd the page).
+   */
+  async function loadGenreRows(libraryId: string) {
+    try {
+      const repo = auth.getRepository();
+      const genres = await repo.getGenres(libraryId);
+      if (genres.length === 0) return;
+
+      const probed = genres.slice(0, MAX_GENRES_PROBED);
+      const rows = await Promise.all(
+        probed.map(async (genre): Promise<GenreRow> => {
+          try {
+            const result = await repo.getItems(libraryId, {
+              includeItemTypes: ["Movie"],
+              genres: [genre.name],
+              sortBy: "CommunityRating",
+              sortOrder: "Descending",
+              recursive: true,
+              limit: SECTION_LIMIT,
+            });
+            return { id: genre.id, name: genre.name, items: result.items };
+          } catch (e) {
+            console.warn(`Failed to load genre row "${genre.name}":`, e);
+            return { id: genre.id, name: genre.name, items: [] };
+          }
+        })
+      );
+
+      const genreRows = rows
+        .filter(row => row.items.length > 0)
+        .sort((a, b) => b.items.length - a.items.length)
+        .slice(0, MAX_GENRE_ROWS);
+
+      update(s => ({ ...s, genreRows }));
+    } catch (e) {
+      console.warn("Failed to load movie genre rows:", e);
+    }
+  }
+
+  function reset() {
+    set(initialState);
+  }
+
+  return {
+    subscribe,
+    loadSections,
+    reset,
+  };
+}
+
+export const movies = createMoviesStore();
+
+export const moviesHeroItems = derived(movies, $m => $m.heroItems);
+export const isMoviesLoading = derived(movies, $m => $m.isLoading);

src/lib/stores/music.ts

@@ -0,0 +1,134 @@
+// Music library landing page data store.
+// Powers the focused music landing: hero + horizontal sliders.
+// TRACES: UR-007, UR-034 | DR-007, DR-038, DR-039
+import { writable, derived } from "svelte/store";
+import type { MediaItem } from "$lib/api/types";
+import { auth } from "./auth";
+import { excludePodcasts } from "$lib/utils/podcastFilter";
+
+interface MusicState {
+  // Albums grouped from recently played tracks (backend handles grouping).
+  recentlyPlayed: MediaItem[];
+  // Most recently added albums in the music library.
+  newlyAdded: MediaItem[];
+  // User playlists.
+  playlists: MediaItem[];
+  // Albums the user has played but hasn't returned to in a while.
+  rediscover: MediaItem[];
+  // Mix of recently played + rediscover, used for the hero banner.
+  heroItems: MediaItem[];
+  isLoading: boolean;
+  error: string | null;
+}
+
+const SECTION_LIMIT = 16;
+
+function createMusicStore() {
+  const initialState: MusicState = {
+    recentlyPlayed: [],
+    newlyAdded: [],
+    playlists: [],
+    rediscover: [],
+    heroItems: [],
+    isLoading: false,
+    error: null,
+  };
+
+  const { subscribe, set, update } = writable<MusicState>(initialState);
+
+  /**
+   * Build the hero rotation from a mix of recently played and rediscover
+   * albums, interleaved so the banner alternates "fresh in your ears" with
+   * "remember this?". De-duplicates by id and prefers items that have artwork.
+   */
+  function buildHero(recent: MediaItem[], rediscover: MediaItem[]): MediaItem[] {
+    const hasArt = (i: MediaItem) =>
+      !!i.primaryImageTag || !!(i.backdropImageTags && i.backdropImageTags.length > 0);
+
+    const recentPool = recent.filter(hasArt);
+    const rediscoverPool = rediscover.filter(hasArt);
+
+    const result: MediaItem[] = [];
+    const seen = new Set<string>();
+    const maxLen = Math.max(recentPool.length, rediscoverPool.length);
+
+    for (let i = 0; i < maxLen && result.length < 6; i++) {
+      for (const candidate of [recentPool[i], rediscoverPool[i]]) {
+        if (candidate && !seen.has(candidate.id)) {
+          seen.add(candidate.id);
+          result.push(candidate);
+        }
+      }
+    }
+
+    return result.slice(0, 6);
+  }
+
+  async function loadSections(libraryId: string) {
+    update(s => ({
+      ...s,
+      isLoading: s.recentlyPlayed.length === 0 && s.newlyAdded.length === 0,
+      error: null,
+    }));
+
+    try {
+      const repo = auth.getRepository();
+
+      const [recentlyPlayed, newlyAdded, playlistsResult, rediscover] = await Promise.all([
+        repo.getRecentlyPlayedAudio(SECTION_LIMIT),
+        repo.getItems(libraryId, {
+          includeItemTypes: ["MusicAlbum"],
+          sortBy: "DateCreated",
+          sortOrder: "Descending",
+          recursive: true,
+          limit: SECTION_LIMIT,
+        }),
+        repo.getItems(libraryId, {
+          includeItemTypes: ["Playlist"],
+          sortBy: "SortName",
+          sortOrder: "Ascending",
+          recursive: true,
+          limit: SECTION_LIMIT,
+        }),
+        repo.getRediscoverAlbums(libraryId, SECTION_LIMIT),
+      ]);
+
+      // HACK: drop the "Podcasts" folder that lives inside the music library.
+      const recentlyPlayedAlbums = excludePodcasts(recentlyPlayed);
+      const newlyAddedAlbums = excludePodcasts(newlyAdded.items);
+      const playlistItems = excludePodcasts(playlistsResult.items);
+      const rediscoverAlbums = excludePodcasts(rediscover);
+
+      const heroItems = buildHero(recentlyPlayedAlbums, rediscoverAlbums);
+
+      update(s => ({
+        ...s,
+        recentlyPlayed: recentlyPlayedAlbums,
+        newlyAdded: newlyAddedAlbums,
+        playlists: playlistItems,
+        rediscover: rediscoverAlbums,
+        heroItems,
+        isLoading: false,
+      }));
+    } catch (error) {
+      const message = error instanceof Error ? error.message : "Failed to load music sections";
+      update(s => ({ ...s, isLoading: false, error: message }));
+      console.error("Failed to load music sections:", error);
+    }
+  }
+
+  function reset() {
+    set(initialState);
+  }
+
+  return {
+    subscribe,
+    loadSections,
+    reset,
+  };
+}
+
+export const music = createMusicStore();
+
+export const musicHeroItems = derived(music, $m => $m.heroItems);
+export const isMusicLoading = derived(music, $m => $m.isLoading);

src/lib/stores/tv.ts

@@ -0,0 +1,176 @@
+// TV library landing page data store.
+// Powers the focused TV landing: hero + horizontal sliders.
+// TRACES: UR-007, UR-023, UR-034 | DR-007, DR-038, DR-039
+import { writable, derived } from "svelte/store";
+import type { MediaItem } from "$lib/api/types";
+import { auth } from "./auth";
+
+/** A single "by genre" row: the genre name plus the series in it. */
+export interface GenreRow {
+  id: string;
+  name: string;
+  items: MediaItem[];
+}
+
+interface TvState {
+  // Episodes the user can resume (continue watching).
+  continueWatching: MediaItem[];
+  // Next unwatched episode per in-progress series.
+  nextUp: MediaItem[];
+  // Recently added items in the TV library (series/seasons/episodes).
+  recentlyAdded: MediaItem[];
+  // One slider per genre (top genres by show count).
+  genreRows: GenreRow[];
+  // Mix used for the hero banner.
+  heroItems: MediaItem[];
+  isLoading: boolean;
+  error: string | null;
+}
+
+const SECTION_LIMIT = 16;
+// How many genre sliders to show, and how many genres to probe to find them.
+const MAX_GENRE_ROWS = 8;
+const MAX_GENRES_PROBED = 20;
+
+function createTvStore() {
+  const initialState: TvState = {
+    continueWatching: [],
+    nextUp: [],
+    recentlyAdded: [],
+    genreRows: [],
+    heroItems: [],
+    isLoading: false,
+    error: null,
+  };
+
+  const { subscribe, set, update } = writable<TvState>(initialState);
+
+  /**
+   * Build the hero rotation. Prefer in-progress episodes (most personal),
+   * then fall back to next-up, then recently added. De-duplicates by id and
+   * prefers items that carry backdrop/primary artwork for a good banner.
+   */
+  function buildHero(
+    continueWatching: MediaItem[],
+    nextUp: MediaItem[],
+    recentlyAdded: MediaItem[]
+  ): MediaItem[] {
+    const hasArt = (i: MediaItem) =>
+      !!(i.backdropImageTags && i.backdropImageTags.length > 0) ||
+      !!(i.parentBackdropImageTags && i.parentBackdropImageTags.length > 0) ||
+      !!i.primaryImageTag;
+
+    const result: MediaItem[] = [];
+    const seen = new Set<string>();
+
+    for (const pool of [continueWatching, nextUp, recentlyAdded]) {
+      for (const item of pool) {
+        if (result.length >= 6) break;
+        if (hasArt(item) && !seen.has(item.id)) {
+          seen.add(item.id);
+          result.push(item);
+        }
+      }
+    }
+
+    return result.slice(0, 6);
+  }
+
+  async function loadSections(libraryId: string) {
+    update(s => ({
+      ...s,
+      isLoading: s.continueWatching.length === 0 && s.recentlyAdded.length === 0,
+      error: null,
+    }));
+
+    try {
+      const repo = auth.getRepository();
+
+      const [resume, nextUp, latest] = await Promise.all([
+        repo.getResumeItems(libraryId, SECTION_LIMIT),
+        repo.getNextUpEpisodes(undefined, SECTION_LIMIT),
+        repo.getLatestItems(libraryId, SECTION_LIMIT),
+      ]);
+
+      // Resume items are already video-only from the server, but keep episodes
+      // (and the occasional movie that lives in a mixed library) defensively.
+      const continueWatching = resume.filter(i => i.type === "Episode" || i.type === "Movie");
+
+      const heroItems = buildHero(continueWatching, nextUp, latest);
+
+      update(s => ({
+        ...s,
+        continueWatching,
+        nextUp,
+        recentlyAdded: latest,
+        heroItems,
+        isLoading: false,
+      }));
+
+      // Genre rows are secondary — load them after the main sections paint so
+      // the page isn't blocked on N per-genre queries.
+      loadGenreRows(libraryId);
+    } catch (error) {
+      const message = error instanceof Error ? error.message : "Failed to load TV sections";
+      update(s => ({ ...s, isLoading: false, error: message }));
+      console.error("Failed to load TV sections:", error);
+    }
+  }
+
+  /**
+   * Build one slider per genre, showing the top-rated series in each. We probe
+   * a bounded set of genres in parallel, drop empty ones, then keep the genres
+   * with the most shows (so niche/near-empty genres don't crowd the page).
+   */
+  async function loadGenreRows(libraryId: string) {
+    try {
+      const repo = auth.getRepository();
+      const genres = await repo.getGenres(libraryId);
+      if (genres.length === 0) return;
+
+      const probed = genres.slice(0, MAX_GENRES_PROBED);
+      const rows = await Promise.all(
+        probed.map(async (genre): Promise<GenreRow> => {
+          try {
+            const result = await repo.getItems(libraryId, {
+              includeItemTypes: ["Series"],
+              genres: [genre.name],
+              sortBy: "CommunityRating",
+              sortOrder: "Descending",
+              recursive: true,
+              limit: SECTION_LIMIT,
+            });
+            return { id: genre.id, name: genre.name, items: result.items };
+          } catch (e) {
+            console.warn(`Failed to load genre row "${genre.name}":`, e);
+            return { id: genre.id, name: genre.name, items: [] };
+          }
+        })
+      );
+
+      const genreRows = rows
+        .filter(row => row.items.length > 0)
+        .sort((a, b) => b.items.length - a.items.length)
+        .slice(0, MAX_GENRE_ROWS);
+
+      update(s => ({ ...s, genreRows }));
+    } catch (e) {
+      console.warn("Failed to load TV genre rows:", e);
+    }
+  }
+
+  function reset() {
+    set(initialState);
+  }
+
+  return {
+    subscribe,
+    loadSections,
+    reset,
+  };
+}
+
+export const tv = createTvStore();
+
+export const tvHeroItems = derived(tv, $t => $t.heroItems);
+export const isTvLoading = derived(tv, $t => $t.isLoading);

src/lib/utils/podcastFilter.ts

@@ -0,0 +1,30 @@
+// HACK: hide "Podcasts" from the music library.
+//
+// The user stores podcasts inside the music library under a folder/album named
+// "Podcasts", so they leak into album/artist/track/playlist queries. Jellyfin's
+// item queries here don't give us a clean server-side exclusion for that folder,
+// so we filter client-side by name. This is intentionally a blunt instrument:
+// anything whose own name, album, or (album) artist is literally "Podcasts" is
+// dropped. If the folder is ever renamed, update PODCAST_FOLDER_NAME.
+import type { MediaItem } from "$lib/api/types";
+
+const PODCAST_FOLDER_NAME = "podcasts";
+
+function isPodcastName(value: string | null | undefined): boolean {
+  return value?.trim().toLowerCase() === PODCAST_FOLDER_NAME;
+}
+
+/** True when an item belongs to the "Podcasts" folder/album and should be hidden. */
+export function isPodcastItem(item: MediaItem): boolean {
+  return (
+    isPodcastName(item.name) ||
+    isPodcastName(item.albumName) ||
+    isPodcastName(item.albumArtist) ||
+    (item.artists?.some(isPodcastName) ?? false)
+  );
+}
+
+/** Remove "Podcasts" entries from a list of music items. */
+export function excludePodcasts(items: MediaItem[]): MediaItem[] {
+  return items.filter((item) => !isPodcastItem(item));
+}