Hotfix: Redis cache clearing, seek drag lag, session persistence (v1.3.4)

Critical Fixes:
- Docker permissions for PostgreSQL/Redis bind mounts
  Fixes #59, fixes #62
- Audio analyzer memory consumption and OOM crashes
  Fixes #21, fixes #26, fixes #53
- LastFM array normalization preventing .map crashes
  Fixes #37, fixes #39
- Wikidata 403 errors from missing User-Agent
  Fixes #57
- Singles directory creation race conditions
  Fixes #58
- Firefox FLAC playback stopping at ~4:34 mark
  Fixes #42, fixes #17

Quality of Life:
- Add Releases link to desktop sidebar navigation
  Fixes #41
- iPhone safe area insets for Dynamic Island/notch
  Fixes #54

Contributors: @arsaboo, @rustyricky, @RustyJonez, @tombatossals

No regressions detected, backward compatible, production ready.

.github/workflows/docker-publish.yml

@@ -1,4 +1,4 @@
-name: Build and Publish Docker Image
+name: Release ${{ github.ref_name }}
 
 on:
     push:

.rooignore

@@ -0,0 +1,279 @@
+# ==============================================================================
+
+# .rooignore - Custom for Lidify (Based on Context Analysis)
+
+# ==============================================================================
+
+# Created: 2026-01-09
+
+# Current token usage: ~177,000 tokens per request
+
+# Target: ~60,000-80,000 tokens per request (60% reduction)
+
+# Expected savings: $335-395/month
+
+# ==============================================================================
+
+# ==============================================================================
+
+# TEST ARTIFACTS - BIGGEST BLOAT (1.4MB found in your project)
+
+# ==============================================================================
+
+# Playwright test reports and results - these are generated artifacts
+
+playwright-report/
+test-results/
+frontend/playwright-report/
+frontend/test-results/
+
+# Test files themselves
+
+_.test.ts
+_.test.tsx
+_.test.js
+_.test.jsx
+_.spec.ts
+_.spec.tsx
+_.spec.js
+_.spec.jsx
+**/**tests**/
+**/tests/
+
+# ==============================================================================
+
+# CONTEXT_PORTAL - Your RAG System (1MB of vector DB data)
+
+# ==============================================================================
+
+# This is YOUR context portal - Roo Code doesn't need to read it!
+
+context_portal/
+context_portal/conport_vector_data/
+context_portal/context.db
+\*.sqlite3
+
+# ==============================================================================
+
+# BUILD ARTIFACTS & CACHES (.next/ = 24MB)
+
+# ==============================================================================
+
+.next/
+dist/
+build/
+out/
+\*.tsbuildinfo
+.turbo/
+
+# ==============================================================================
+
+# DEPENDENCIES - Never needed (429M backend + 729M frontend)
+
+# ==============================================================================
+
+node_modules/
+.pnp
+.pnp.js
+.yarn/
+
+# Lock files (488KB total)
+
+package-lock.json
+yarn.lock
+pnpm-lock.yaml
+**/node_modules/**/yarn.lock
+
+# ==============================================================================
+
+# IMAGES & MEDIA - (3MB+ of screenshots)
+
+# ==============================================================================
+
+# All image formats
+
+_.png
+_.jpg
+_.jpeg
+_.gif
+_.webp
+_.svg
+_.ico
+_.bmp
+
+# Specifically your screenshot directories
+
+assets/screenshots/
+frontend/assets/splash.png
+frontend/assets/splash-dark.png
+
+# ==============================================================================
+
+# DOCS - Large deployment doc (312KB)
+
+# ==============================================================================
+
+# Keep README.md, CONTRIBUTING.md, CHANGELOG.md
+
+# Exclude large pending deploy docs
+
+docs/PENDING_DEPLOY-1.md
+
+# ==============================================================================
+
+# DATABASE MIGRATIONS - Keep recent, exclude none (all are 2025+)
+
+# ==============================================================================
+
+# Your migrations are all from 2025/2026, so keep them all
+
+# If you add older migrations later:
+
+# backend/prisma/migrations/2024\*/
+
+# ==============================================================================
+
+# VERSION CONTROL
+
+# ==============================================================================
+
+.git/
+.gitignore
+.gitattributes
+
+# ==============================================================================
+
+# SOULARR - External project (separate tool)
+
+# ==============================================================================
+
+# If this is a separate tool/subproject, exclude it
+
+soularr/
+
+# ==============================================================================
+
+# IDE & EDITOR
+
+# ==============================================================================
+
+.vscode/
+.idea/
+\*.sublime-workspace
+.DS_Store
+desktop.ini
+
+# Roo-specific directories (don't need to analyze Roo's own metadata)
+
+.roo/
+.claude/
+
+# ==============================================================================
+
+# LOGS & TEMP
+
+# ==============================================================================
+
+_.log
+npm-debug.log_
+yarn-debug.log*
+yarn-error.log*
+_.tmp
+_.temp
+tmp/
+temp/
+
+# ==============================================================================
+
+# ENVIRONMENT FILES
+
+# ==============================================================================
+
+.env.local
+.env.\*.local
+.env.production
+.env.development
+
+# ==============================================================================
+
+# DOCKER (Keep these - you modify them)
+
+# ==============================================================================
+
+# Keeping Docker files as you have 8 docker-compose files
+
+# Uncomment if you rarely modify:
+
+# Dockerfile
+
+# docker-compose\*.yml
+
+# ==============================================================================
+
+# GITHUB WORKFLOWS
+
+# ==============================================================================
+
+.github/workflows/
+
+# ==============================================================================
+
+# PYTHON CACHE (from services/audio-analyzer)
+
+# ==============================================================================
+
+**pycache**/
+_.pyc
+_.pyo
+\*.pyd
+.Python
+
+# ==============================================================================
+
+# VERIFICATION CHECKLIST
+
+# ==============================================================================
+
+# After adding this file:
+
+#
+
+# 1. Restart Roo Code
+
+# 2. Make a simple request (e.g., "explain backend/src/routes/library.ts")
+
+# 3. Check OpenRouter activity: https://openrouter.ai/activity
+
+# 4. Verify token count: Should be ~60K-80K (down from 177K)
+
+#
+
+# If still high:
+
+# - Check if node_modules/ is truly excluded
+
+# - Verify .next/ is excluded
+
+# - Check if test files are still being loaded
+
+#
+
+# If too aggressive (AI can't find files):
+
+# - Remove specific exclusions one at a time
+
+# - Start by uncommenting Docker files
+
+# - Then uncomment docs if needed
+
+#
+
+# Expected cost per request:
+
+# - Before: $0.141 (177K tokens)
+
+# - After: $0.055-0.070 (60-80K tokens)
+
+# - Savings: 50-60% reduction
+
+# ==============================================================================

CHANGELOG.md

@@ -5,6 +5,77 @@ All notable changes to Lidify will be documented in this file.
 The format is based on [Keep a Changelog](https://keepachangelog.com/en/1.0.0/),
 and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0.html).
 
+## [1.3.3] - 2025-01-07
+
+Bug fix patch release addressing 6 P1 critical issues and 2 P2 quality-of-life improvements.
+
+### Fixed
+
+#### Critical (P1)
+-   **Docker:** PostgreSQL/Redis bind mount permission errors on Linux hosts ([#59](https://github.com/Chevron7Locked/lidify/issues/59)) - @arsaboo via [#62](https://github.com/Chevron7Locked/lidify/pull/62)
+-   **Audio Analyzer:** Memory consumption/OOM crashes with large libraries ([#21](https://github.com/Chevron7Locked/lidify/issues/21), [#26](https://github.com/Chevron7Locked/lidify/issues/26)) - @rustyricky via [#53](https://github.com/Chevron7Locked/lidify/pull/53)
+-   **LastFM:** ".map is not a function" crashes with obscure artists ([#37](https://github.com/Chevron7Locked/lidify/issues/37)) - @RustyJonez via [#39](https://github.com/Chevron7Locked/lidify/pull/39)
+-   **Wikidata:** 403 Forbidden errors from missing User-Agent header ([#57](https://github.com/Chevron7Locked/lidify/issues/57))
+-   **Downloads:** Singles directory creation race conditions ([#58](https://github.com/Chevron7Locked/lidify/issues/58))
+-   **Firefox:** FLAC playback stopping at ~4:34 mark on large files ([#42](https://github.com/Chevron7Locked/lidify/issues/42), [#17](https://github.com/Chevron7Locked/lidify/issues/17))
+
+#### Quality of Life (P2)
+-   **Desktop UI:** Added missing "Releases" link to desktop sidebar navigation ([#41](https://github.com/Chevron7Locked/lidify/issues/41))
+-   **iPhone:** Dynamic Island/notch overlapping TopBar buttons ([#54](https://github.com/Chevron7Locked/lidify/issues/54))
+
+### Technical Details
+
+-   **Docker Permissions (#62):** Creates `/data/postgres` and `/data/redis` directories with proper ownership; validates write permissions at startup using `gosu <user> test -w`
+-   **Audio Analyzer Memory (#53):** TensorFlow GPU memory growth enabled; `MAX_ANALYZE_SECONDS` configurable (default 90s); explicit garbage collection in finally blocks
+-   **LastFM Normalization (#39):** `normalizeToArray()` utility wraps single-object API responses; protects 5 locations in artist discovery endpoints
+-   **Wikidata User-Agent (#57):** All 4 API endpoints now use configured axios client with proper User-Agent header
+-   **Singles Directory (#58):** Replaced TOCTOU `existsSync()`+`mkdirSync()` pattern with idempotent `mkdir({recursive: true})`
+-   **Firefox FLAC (#42):** Replaced Express `res.sendFile()` with manual range request handling via `fs.createReadStream()` with proper `Content-Range` headers
+-   **Desktop Releases (#41):** Single-line addition to Sidebar.tsx navigation array
+-   **iPhone Safe Area (#54):** TopBar and AuthenticatedLayout use `env(safe-area-inset-top)` CSS environment variable
+
+### Deferred to Future Release
+
+-   **PR #49** - Playlist visibility toggle (needs PR review)
+-   **PR #47** - Mood bucket tags (already implemented, verify and close)
+-   **PR #36** - Docker --user flag (needs security review)
+
+### Contributors
+
+Thanks to everyone who contributed to this release:
+
+-   @arsaboo - Docker bind mount permissions fix ([#62](https://github.com/Chevron7Locked/lidify/pull/62))
+-   @rustyricky - Audio analyzer memory limits ([#53](https://github.com/Chevron7Locked/lidify/pull/53))
+-   @RustyJonez - LastFM array normalization ([#39](https://github.com/Chevron7Locked/lidify/pull/39))
+-   @tombatossals - Testing and validation
+
+---
+
+## [1.3.2] - 2025-01-07
+
+### Fixed
+- Mobile scrolling blocked by pull-to-refresh component
+- Pull-to-refresh component temporarily disabled (will be properly fixed in v1.4)
+
+### Technical Details
+- Root cause: CSS flex chain break (`h-full`) and touch event interference
+- Implemented early return to bypass problematic wrapper while preserving child rendering
+- TODO: Re-enable in v1.4 with proper CSS fix (`flex-1 flex flex-col min-h-0`)
+
+## [1.3.1] - 2025-01-07
+
+### Fixed
+- Production database schema mismatch causing SystemSettings endpoints to fail
+- Added missing `downloadSource` and `primaryFailureFallback` columns to SystemSettings table
+
+### Database Migrations
+- `20260107000000_add_download_source_columns` - Idempotent migration adds missing columns with defaults
+
+### Technical Details
+- Root cause: Migration gap between squashed init migration and production database setup
+- Uses PostgreSQL IF NOT EXISTS pattern for safe deployment across all environments
+- Default values: `downloadSource='soulseek'`, `primaryFailureFallback='none'`
+
 ## [1.3.0] - 2026-01-06
 
 ### Added

Dockerfile

@@ -193,6 +193,7 @@ priority=10
 
 [program:redis]
 command=/usr/bin/redis-server --dir /data/redis --appendonly yes
+user=redis
 autostart=true
 autorestart=true
 stdout_logfile=/dev/stdout
@@ -235,7 +236,7 @@ stdout_logfile=/dev/stdout
 stdout_logfile_maxbytes=0
 stderr_logfile=/dev/stderr
 stderr_logfile_maxbytes=0
-environment=DATABASE_URL="postgresql://lidify:lidify@localhost:5432/lidify",REDIS_URL="redis://localhost:6379",MUSIC_PATH="/music",BATCH_SIZE="10",SLEEP_INTERVAL="5"
+environment=DATABASE_URL="postgresql://lidify:lidify@localhost:5432/lidify",REDIS_URL="redis://localhost:6379",MUSIC_PATH="/music",BATCH_SIZE="10",SLEEP_INTERVAL="5",MAX_ANALYZE_SECONDS="90"
 priority=50
 EOF
 
@@ -274,10 +275,33 @@ if [ -z "$PG_BIN" ]; then
 fi
 echo "Using PostgreSQL from: $PG_BIN"
 
-# Fix permissions on data directories (may have different UID from previous container)
+# Prepare data directories (bind-mount safe)
-echo "Fixing data directory permissions..."
+echo "Preparing data directories..."
-chown -R postgres:postgres /data/postgres /run/postgresql 2>/dev/null || true
+mkdir -p /data/postgres /data/redis /run/postgresql
-chmod 700 /data/postgres 2>/dev/null || true
+
+if id postgres >/dev/null 2>&1; then
+    chown -R postgres:postgres /data/postgres /run/postgresql 2>/dev/null || true
+    chmod 700 /data/postgres 2>/dev/null || true
+    if ! gosu postgres test -w /data/postgres; then
+        POSTGRES_UID=$(id -u postgres)
+        POSTGRES_GID=$(id -g postgres)
+        echo "ERROR: /data/postgres is not writable by postgres (${POSTGRES_UID}:${POSTGRES_GID})."
+        echo "If you bind-mount /data, ensure the host path is writable by that UID/GID."
+        exit 1
+    fi
+fi
+
+if id redis >/dev/null 2>&1; then
+    chown -R redis:redis /data/redis 2>/dev/null || true
+    chmod 700 /data/redis 2>/dev/null || true
+    if ! gosu redis test -w /data/redis; then
+        REDIS_UID=$(id -u redis)
+        REDIS_GID=$(id -g redis)
+        echo "ERROR: /data/redis is not writable by redis (${REDIS_UID}:${REDIS_GID})."
+        echo "If you bind-mount /data, ensure the host path is writable by that UID/GID."
+        exit 1
+    fi
+fi
 
 # Clean up stale PID file if exists
 rm -f /data/postgres/postmaster.pid 2>/dev/null || true

README.md

@@ -14,7 +14,7 @@ Lidify is built for music lovers who want the convenience of streaming services
 
 ## A Note on Native Apps
 
-I got a little and PWA are the priority. Once the core experience is solid and properly tested, a native mobile app (likely React Native) is on the roadmap. The PWA works great for most cases for now.
+Once the core experience is solid and properly tested, a native mobile app (likely React Native) is on the roadmap. The PWA works great for most cases for now.
 
 Thanks for your patience while I work through this.
 
@@ -45,6 +45,7 @@ Thanks for your patience while I work through this.
 -   **Stream your library** - FLAC, MP3, AAC, OGG, and other common formats work out of the box
 -   **Automatic cataloging** - Lidify scans your library and enriches it with metadata from MusicBrainz and Last.fm
 -   **Audio transcoding** - Stream at original quality or transcode on-the-fly (320kbps, 192kbps, or 128kbps)
+-   **Ultra-wide support** - Library grid scales up to 8 columns on large displays
 
 <p align="center">
   <img src="assets/screenshots/desktop-library.png" alt="Library View" width="800">
@@ -66,6 +67,8 @@ Thanks for your patience while I work through this.
     -   Dynamic genre and decade stations generated from your library
 -   **Discover Weekly** - Weekly playlists of new music tailored to your listening habits (requires Lidarr)
 -   **Artist recommendations** - Find similar artists based on what you already love
+-   **Artist name resolution** - Smart alias lookup via Last.fm (e.g., "of mice" → "Of Mice & Men")
+-   **Discography sorting** - Sort artist albums by year or date added
 -   **Deezer previews** - Preview tracks you don't own before adding them to your library
 -   **Vibe matching** - Find tracks that match your current mood (see [The Vibe System](#the-vibe-system))
 
@@ -74,6 +77,7 @@ Thanks for your patience while I work through this.
 -   **Subscribe via RSS** - Search iTunes for podcasts and subscribe directly
 -   **Track progress** - Pick up where you left off across devices
 -   **Episode management** - Browse episodes, mark as played, and manage your subscriptions
+-   **Mobile skip buttons** - Jump ±30 seconds on mobile for easy navigation
 
 <p align="center">
   <img src="assets/screenshots/desktop-podcasts.png" alt="Podcasts" width="800">
@@ -84,6 +88,7 @@ Thanks for your patience while I work through this.
 -   **Audiobookshelf integration** - Connect your existing Audiobookshelf instance
 -   **Unified experience** - Browse and listen to audiobooks alongside your music
 -   **Progress sync** - Your listening position syncs with Audiobookshelf
+-   **Mobile skip buttons** - Jump ±30 seconds on mobile for easy chapter navigation
 
 <p align="center">
   <img src="assets/screenshots/desktop-audiobooks.png" alt="Audiobooks" width="800">
@@ -172,7 +177,7 @@ Lidify works as a PWA on mobile devices, giving you a native app-like experience
 
 -   Full streaming functionality
 -   Background audio playback
--   Lock screen / notification media controls (via Media Session API)
+-   Lock screen and notification media controls (iOS Control Center and Android notifications)
 -   Offline caching for faster loads
 -   Installable icon on home screen
 
@@ -270,6 +275,16 @@ docker compose pull
 docker compose up -d
 ```
 
+### Bind-mounting `/data` on Linux
+
+Named volumes are recommended. If you bind-mount `/data`, make sure required subdirectories exist and are writable by the container service users.
+
+```bash
+mkdir -p /path/to/lidify-data/postgres /path/to/lidify-data/redis
+```
+
+If startup logs report a permission error, `chown` the host path to the UID/GID shown in the logs (for example, the postgres user).
+
 ---
 
 Lidify will begin scanning your music library automatically. Depending on the size of your collection, this may take a few minutes to several hours.
@@ -310,12 +325,17 @@ docker pull chevron7locked/lidify:nightly
 
 The unified Lidify container handles most configuration automatically. Here are the available options:
 
-| Variable              | Default                            | Description                                                                 |
+| Variable                            | Default                            | Description                                                                 |
-| --------------------- | ---------------------------------- | --------------------------------------------------------------------------- |
+| ----------------------------------- | ---------------------------------- | --------------------------------------------------------------------------- |
-| `SESSION_SECRET`      | Auto-generated                     | Session encryption key (recommended to set for persistence across restarts) |
+| `SESSION_SECRET`                    | Auto-generated                     | Session encryption key (recommended to set for persistence across restarts) |
-| `TZ`                  | `UTC`                              | Timezone for the container                                                  |
+| `SETTINGS_ENCRYPTION_KEY`           | Required                           | Encryption key for stored credentials (generate with `openssl rand -base64 32`) |
-| `LIDIFY_CALLBACK_URL` | `http://host.docker.internal:3030` | URL for Lidarr webhook callbacks (see [Lidarr integration](#lidarr))        |
+| `TZ`                                | `UTC`                              | Timezone for the container                                                  |
-| `NUM_WORKERS`         | `2`                                | Number of parallel workers for audio analysis                               |
+| `PORT`                              | `3030`                             | Port to access Lidify                                                       |
+| `LIDIFY_CALLBACK_URL`               | `http://host.docker.internal:3030` | URL for Lidarr webhook callbacks (see [Lidarr integration](#lidarr))        |
+| `AUDIO_ANALYSIS_WORKERS`            | `2`                                | Number of parallel workers for audio analysis (1-8)                         |
+| `AUDIO_ANALYSIS_THREADS_PER_WORKER` | `1`                                | Threads per worker for TensorFlow/FFT operations (1-4)                      |
+| `LOG_LEVEL`                         | `warn` (prod) / `debug` (dev)      | Logging verbosity: debug, info, warn, error, silent                         |
+| `DOCS_PUBLIC`                       | `false`                            | Set to `true` to allow public access to API docs in production              |
 
 The music library path is configured via Docker volume mount (`-v /path/to/music:/music`).
 
@@ -344,7 +364,7 @@ Lidify uses several sensitive environment variables. Never commit your `.env` fi
 | Variable                  | Purpose                        | Required          |
 | ------------------------- | ------------------------------ | ----------------- |
 | `SESSION_SECRET`          | Session encryption (32+ chars) | Yes               |
-| `SETTINGS_ENCRYPTION_KEY` | Encrypts stored credentials    | Recommended       |
+| `SETTINGS_ENCRYPTION_KEY` | Encrypts stored credentials    | Yes               |
 | `SOULSEEK_USERNAME`       | Soulseek login                 | If using Soulseek |
 | `SOULSEEK_PASSWORD`       | Soulseek password              | If using Soulseek |
 | `LIDARR_API_KEY`          | Lidarr integration             | If using Lidarr   |
@@ -352,6 +372,24 @@ Lidify uses several sensitive environment variables. Never commit your `.env` fi
 | `LASTFM_API_KEY`          | Artist recommendations         | Optional          |
 | `FANART_API_KEY`          | Artist images                  | Optional          |
 
+### Authentication & Session Security
+
+-   **JWT tokens** - Access tokens expire after 24 hours; refresh tokens after 30 days
+-   **Token refresh** - Automatic token refresh via `/api/auth/refresh` endpoint
+-   **Password changes** - Changing your password invalidates all existing sessions
+-   **Session cookies** - Secured with `httpOnly`, `sameSite=strict`, and `secure` (in production)
+-   **Encryption validation** - Encryption key is validated on startup to prevent insecure defaults
+
+### Webhook Security
+
+-   **Lidarr webhooks** - Support signature verification with configurable secret
+-   Configure the webhook secret in Settings → Lidarr for additional security
+
+### Admin Dashboard Security
+
+-   **Bull Board** - Job queue dashboard at `/admin/queues` requires authenticated admin user
+-   **API Documentation** - Swagger docs at `/api-docs` require authentication in production (unless `DOCS_PUBLIC=true`)
+
 ### VPN Configuration (Optional)
 
 If using Mullvad VPN for Soulseek:
@@ -367,7 +405,7 @@ If using Mullvad VPN for Soulseek:
 openssl rand -base64 32
 
 # Generate encryption key
-openssl rand -hex 32
+openssl rand -base64 32
 ```
 
 ### Network Security
@@ -389,7 +427,7 @@ Connect Lidify to your Lidarr instance to request and download new music directl
 **What you get:**
 
 -   Browse artists and albums you don't own
--               Request downloads with a single click
+-                 Request downloads with a single click
 -   Discover Weekly playlists that automatically download new recommendations
 -   Automatic library sync when Lidarr finishes importing
 
@@ -601,6 +639,22 @@ Administrators have access to additional settings:
 -   **Cache Management** - Clear caches if needed
 -   **Advanced** - Download retry settings, concurrent download limits
 
+### Download Settings
+
+Configure how Lidify acquires new music in Settings → Downloads:
+
+-   **Primary Source** - Choose between Soulseek or Lidarr as your main download source
+-   **Fallback Behavior** - Optionally fall back to the other source if the primary fails
+-   **Stale Job Cleanup** - Clear stuck Discovery batches and downloads that aren't progressing
+
+### Enrichment Settings
+
+Control metadata enrichment in Settings → Cache & Automation:
+
+-   **Enrichment Speed** - Adjust concurrency (1-5x) to balance speed vs. system load
+-   **Failure Notifications** - Get notified when enrichment fails for specific items
+-   **Retry/Skip Modal** - Choose to retry failed items or skip them to continue processing
+
 ### Activity Panel
 
 The Activity Panel provides real-time visibility into downloads and system events:
@@ -619,7 +673,16 @@ For programmatic access to Lidify:
 2. Generate a new key with a descriptive name
 3. Use the key in the `Authorization` header: `Bearer YOUR_API_KEY`
 
-API documentation is available at `/api-docs` when the backend is running.
+API documentation is available at `/api-docs` when the backend is running (requires authentication in production).
+
+### Bull Board Dashboard
+
+Monitor background job queues at `/admin/queues`:
+
+-   View active, waiting, completed, and failed jobs
+-   Retry or remove stuck jobs
+-   Monitor download progress and enrichment tasks
+-   Requires admin authentication
 
 ---
 

analyze-context-bloat-v2.sh

@@ -0,0 +1,200 @@
+#!/bin/bash
+
+# ==============================================================================
+# analyze-context-bloat.sh
+# ==============================================================================
+# Purpose: Find large files in your project that are bloating Roo Code context
+# Usage: Run this in your Lidify project root directory
+#        chmod +x analyze-context-bloat.sh && ./analyze-context-bloat.sh
+# ==============================================================================
+
+echo "=============================================================================="
+echo "Lidify Context Bloat Analysis"
+echo "=============================================================================="
+echo ""
+echo "Analyzing your project to find files that should be excluded from Roo Code..."
+echo ""
+
+# Check if we're in a project directory (monorepo structure)
+if [ ! -f "backend/package.json" ] && [ ! -f "frontend/package.json" ] && [ ! -f "package.json" ]; then
+    echo "❌ Error: Run this script from your Lidify project root directory"
+    echo "   (Looking for backend/package.json or frontend/package.json or package.json)"
+    exit 1
+fi
+
+echo "✅ Found project structure (monorepo detected)"
+echo ""
+
+echo "📊 TOP 30 LARGEST FILES (excluding node_modules):"
+echo "=============================================================================="
+find . -type f -not -path "*/node_modules/*" -not -path "*/.git/*" -not -path "*/.next/*" -not -path "*/dist/*" -exec du -h {} + 2>/dev/null | sort -rh | head -30
+echo ""
+
+echo "📦 DIRECTORY SIZES (top-level):"
+echo "=============================================================================="
+du -h --max-depth=1 . 2>/dev/null | sort -rh
+echo ""
+
+echo "📦 SUBDIRECTORY SIZES (backend, frontend, services):"
+echo "=============================================================================="
+for dir in backend frontend services scripts; do
+    if [ -d "$dir" ]; then
+        echo ""
+        echo "--- $dir/ ---"
+        du -h --max-depth=2 "$dir" 2>/dev/null | sort -rh | head -10
+    fi
+done
+echo ""
+
+echo "🖼️  IMAGE FILES (all types):"
+echo "=============================================================================="
+find . -type f \( -name "*.png" -o -name "*.jpg" -o -name "*.jpeg" -o -name "*.gif" -o -name "*.webp" -o -name "*.svg" -o -name "*.ico" \) -not -path "*/node_modules/*" 2>/dev/null | wc -l
+echo "Total image files found"
+echo ""
+echo "Largest images:"
+find . -type f \( -name "*.png" -o -name "*.jpg" -o -name "*.jpeg" -o -name "*.gif" -o -name "*.webp" \) -not -path "*/node_modules/*" -exec du -h {} + 2>/dev/null | sort -rh | head -20
+echo ""
+
+echo "📝 LOCK FILES & GENERATED CODE:"
+echo "=============================================================================="
+find . -type f \( -name "package-lock.json" -o -name "yarn.lock" -o -name "pnpm-lock.yaml" -o -name "*.tsbuildinfo" \) -exec du -h {} \; 2>/dev/null
+echo ""
+
+echo "📜 MIGRATION FILES:"
+echo "=============================================================================="
+if [ -d "backend/prisma/migrations" ]; then
+    echo "Total migration directory size:"
+    du -sh backend/prisma/migrations 2>/dev/null
+    echo ""
+    echo "Number of migrations:"
+    ls -1 backend/prisma/migrations 2>/dev/null | wc -l
+    echo ""
+    echo "Oldest migrations (first 10):"
+    ls -1 backend/prisma/migrations 2>/dev/null | head -10
+    echo ""
+    echo "Newest migrations (last 5):"
+    ls -1 backend/prisma/migrations 2>/dev/null | tail -5
+else
+    echo "No migrations directory found"
+fi
+echo ""
+
+echo "🗂️  FILE TYPE BREAKDOWN:"
+echo "=============================================================================="
+echo "TypeScript/JavaScript files:"
+find . -type f \( -name "*.ts" -o -name "*.tsx" -o -name "*.js" -o -name "*.jsx" \) -not -path "*/node_modules/*" -not -path "*/.next/*" 2>/dev/null | wc -l
+echo ""
+echo "JSON files:"
+find . -type f -name "*.json" -not -path "*/node_modules/*" 2>/dev/null | wc -l
+echo ""
+echo "CSS/Style files:"
+find . -type f \( -name "*.css" -o -name "*.scss" -o -name "*.sass" \) -not -path "*/node_modules/*" 2>/dev/null | wc -l
+echo ""
+echo "Markdown files:"
+find . -type f -name "*.md" 2>/dev/null | wc -l
+echo ""
+
+echo "💾 ESTIMATED TOKEN COUNT:"
+echo "=============================================================================="
+# Rough estimation: 1 token ≈ 4 characters
+total_chars=$(find . -type f \( -name "*.ts" -o -name "*.tsx" -o -name "*.js" -o -name "*.jsx" -o -name "*.json" -o -name "*.md" -o -name "*.css" -o -name "*.yml" -o -name "*.yaml" \) -not -path "*/node_modules/*" -not -path "*/.git/*" -not -path "*/.next/*" -not -path "*/dist/*" -exec cat {} \; 2>/dev/null | wc -c)
+estimated_tokens=$((total_chars / 4))
+
+if [ $estimated_tokens -gt 0 ]; then
+    echo "Total characters in text files: $(printf "%'d" $total_chars)"
+    echo "Estimated current token count: ~$(printf "%'d" $estimated_tokens) tokens"
+    echo ""
+    
+    optimized_tokens=$((estimated_tokens * 40 / 100))
+    echo "Estimated AFTER .rooignore: ~$(printf "%'d" $optimized_tokens) tokens (60% reduction)"
+else
+    echo "Could not calculate token estimate"
+fi
+echo ""
+
+echo "🎯 LARGE FILES ANALYSIS:"
+echo "=============================================================================="
+
+echo "Large JSON files (>50KB):"
+large_json=$(find . -type f -name "*.json" -not -path "*/node_modules/*" -not -name "package.json" -not -name "tsconfig.json" -size +50k 2>/dev/null)
+if [ -n "$large_json" ]; then
+    echo "$large_json" | while read file; do
+        size=$(du -h "$file" | cut -f1)
+        echo "  $size - $file"
+    done
+else
+    echo "  None found"
+fi
+echo ""
+
+echo "Large CSS files (>30KB):"
+large_css=$(find . -type f \( -name "*.css" -o -name "*.scss" \) -not -path "*/node_modules/*" -size +30k 2>/dev/null)
+if [ -n "$large_css" ]; then
+    echo "$large_css" | while read file; do
+        size=$(du -h "$file" | cut -f1)
+        echo "  $size - $file"
+    done
+else
+    echo "  None found"
+fi
+echo ""
+
+echo "Test files:"
+test_count=$(find . -type f \( -name "*.test.*" -o -name "*.spec.*" \) -not -path "*/node_modules/*" 2>/dev/null | wc -l)
+echo "  Found $test_count test files"
+if [ "$test_count" -gt 0 ]; then
+    echo "  Consider excluding with: *.test.* and *.spec.*"
+fi
+echo ""
+
+python_files=$(find . -type f -name "*.py" -not -path "*/node_modules/*" 2>/dev/null | wc -l)
+if [ "$python_files" -gt 0 ]; then
+    echo "Python files (services):"
+    echo "  Found $python_files Python files"
+    echo "  Largest Python files:"
+    find . -type f -name "*.py" -not -path "*/node_modules/*" -exec du -h {} + 2>/dev/null | sort -rh | head -5
+    echo ""
+fi
+
+docker_files=$(find . -maxdepth 2 -type f \( -name "Dockerfile*" -o -name "docker-compose*.yml" \) 2>/dev/null | wc -l)
+if [ "$docker_files" -gt 0 ]; then
+    echo "Docker configuration files:"
+    find . -maxdepth 2 -type f \( -name "Dockerfile*" -o -name "docker-compose*.yml" \) -exec du -h {} \; 2>/dev/null
+    echo ""
+fi
+
+echo "=============================================================================="
+echo "🎯 RECOMMENDED .rooignore ADDITIONS:"
+echo "=============================================================================="
+echo ""
+echo "Based on this analysis, your .rooignore should definitely include:"
+echo ""
+echo "1. node_modules/ (if exists)"
+echo "2. Lock files (package-lock.json, yarn.lock)"
+echo "3. All images in assets/screenshots/"
+echo "4. Build artifacts (.next/, dist/, build/)"
+echo "5. Old migrations (backend/prisma/migrations/2024*/)"
+echo ""
+
+if [ -n "$large_json" ]; then
+    echo "6. Large JSON files:"
+    echo "$large_json" | while read file; do
+        echo "   $file"
+    done
+    echo ""
+fi
+
+if [ "$test_count" -gt 5 ]; then
+    echo "7. Test files (*.test.*, *.spec.*)"
+    echo ""
+fi
+
+echo "=============================================================================="
+echo "✅ Analysis complete!"
+echo ""
+echo "Next steps:"
+echo "1. Share this output with Claude"
+echo "2. Claude will create a custom .rooignore for your project"
+echo "3. Copy .rooignore to project root"
+echo "4. Make a Roo Code request and verify token reduction"
+echo "=============================================================================="

analyze-context-bloat.sh

@@ -0,0 +1,107 @@
+#!/bin/bash
+
+# ==============================================================================
+# analyze-context-bloat.sh
+# ==============================================================================
+# Purpose: Find large files in your project that are bloating Roo Code context
+# Usage: Run this in your Lidify project root directory
+#        chmod +x analyze-context-bloat.sh && ./analyze-context-bloat.sh
+# ==============================================================================
+
+echo "=============================================================================="
+echo "Lidify Context Bloat Analysis"
+echo "=============================================================================="
+echo ""
+echo "Analyzing your project to find files that should be excluded from Roo Code..."
+echo ""
+
+# Check if we're in a project directory
+if [ ! -f "package.json" ]; then
+    echo "❌ Error: Run this script from your Lidify project root directory"
+    exit 1
+fi
+
+echo "📊 TOP 20 LARGEST FILES (excluding node_modules):"
+echo "=============================================================================="
+find . -type f -not -path "*/node_modules/*" -not -path "*/.git/*" -not -path "*/.next/*" -exec du -h {} + 2>/dev/null | sort -rh | head -20
+echo ""
+
+echo "📦 DIRECTORY SIZES (excluding node_modules):"
+echo "=============================================================================="
+du -h --max-depth=2 . 2>/dev/null | grep -v node_modules | sort -rh | head -20
+echo ""
+
+echo "🖼️  IMAGE FILES TAKING UP SPACE:"
+echo "=============================================================================="
+find . -type f \( -name "*.png" -o -name "*.jpg" -o -name "*.jpeg" -o -name "*.gif" -o -name "*.webp" \) -not -path "*/node_modules/*" -exec du -h {} + 2>/dev/null | sort -rh | head -20
+echo ""
+
+echo "📝 LOCK FILES & GENERATED CODE:"
+echo "=============================================================================="
+find . -type f \( -name "package-lock.json" -o -name "yarn.lock" -o -name "pnpm-lock.yaml" -o -name "*.tsbuildinfo" \) -exec du -h {} \;
+echo ""
+
+echo "📜 MIGRATION FILES:"
+echo "=============================================================================="
+if [ -d "backend/prisma/migrations" ]; then
+    echo "Total migration directory size:"
+    du -sh backend/prisma/migrations
+    echo ""
+    echo "Number of migrations:"
+    ls -1 backend/prisma/migrations | wc -l
+    echo ""
+    echo "Oldest migrations (first 5):"
+    ls -1 backend/prisma/migrations | head -5
+    echo ""
+    echo "Newest migrations (last 5):"
+    ls -1 backend/prisma/migrations | tail -5
+else
+    echo "No migrations directory found"
+fi
+echo ""
+
+echo "💾 ESTIMATED TOKEN COUNT:"
+echo "=============================================================================="
+# Rough estimation: 1 token ≈ 4 characters
+total_chars=$(find . -type f -not -path "*/node_modules/*" -not -path "*/.git/*" -not -path "*/.next/*" -name "*.ts" -o -name "*.tsx" -o -name "*.js" -o -name "*.jsx" -o -name "*.json" -o -name "*.md" 2>/dev/null | xargs cat 2>/dev/null | wc -c)
+estimated_tokens=$((total_chars / 4))
+echo "Estimated current token count: ~$(printf "%'d" $estimated_tokens) tokens"
+echo ""
+
+echo "🎯 RECOMMENDED .rooignore ADDITIONS:"
+echo "=============================================================================="
+echo "Based on this analysis, consider adding these to .rooignore:"
+echo ""
+
+# Find large JSON files
+large_json=$(find . -type f -name "*.json" -not -path "*/node_modules/*" -not -name "package.json" -not -name "tsconfig.json" -size +100k -exec du -h {} \; 2>/dev/null)
+if [ -n "$large_json" ]; then
+    echo "Large JSON files (>100KB):"
+    echo "$large_json"
+    echo ""
+fi
+
+# Find CSS/SCSS files if they're large
+large_css=$(find . -type f \( -name "*.css" -o -name "*.scss" \) -not -path "*/node_modules/*" -size +50k -exec du -h {} \; 2>/dev/null)
+if [ -n "$large_css" ]; then
+    echo "Large CSS files (>50KB):"
+    echo "$large_css"
+    echo ""
+fi
+
+# Find test files
+test_files=$(find . -type f \( -name "*.test.*" -o -name "*.spec.*" \) -not -path "*/node_modules/*" | wc -l)
+if [ "$test_files" -gt 0 ]; then
+    echo "Found $test_files test files - consider excluding with: *.test.* and *.spec.*"
+    echo ""
+fi
+
+echo "=============================================================================="
+echo "✅ Analysis complete!"
+echo ""
+echo "Next steps:"
+echo "1. Copy .rooignore to your project root"
+echo "2. Add any large files shown above to .rooignore"
+echo "3. Make a Roo Code request and check token count in OpenRouter"
+echo "4. Target: 60-80K tokens (down from 177K)"
+echo "=============================================================================="

backend/docker-entrypoint.sh

@@ -36,6 +36,17 @@ npx prisma migrate deploy
 echo "[DB] Generating Prisma client..."
 npx prisma generate
 
+# Clear Redis cache on deployment to prevent stale data (e.g., 404 images)
+echo "[REDIS] Clearing cache for fresh deployment..."
+node -e "
+const { createClient } = require('redis');
+const client = createClient({ url: process.env.REDIS_URL || 'redis://redis:6379' });
+client.connect()
+  .then(() => client.flushAll())
+  .then(() => { console.log('[REDIS] Cache cleared successfully'); return client.quit(); })
+  .catch(err => { console.warn('[REDIS] Cache clear failed (non-critical):', err.message); });
+" || echo "[REDIS] Cache clear skipped (Redis unavailable)"
+
 # Generate session secret if not provided
 if [ -z "$SESSION_SECRET" ] || [ "$SESSION_SECRET" = "changeme-generate-secure-key" ]; then
   echo "[WARN] SESSION_SECRET not set or using default. Generating random key..."

backend/prisma/migrations/20260107000000_add_download_source_columns/migration.sql

@@ -0,0 +1,21 @@
+-- Add downloadSource column if it doesn't exist (idempotent)
+DO $$
+BEGIN
+    IF NOT EXISTS (
+        SELECT 1 FROM information_schema.columns
+        WHERE table_name = 'SystemSettings' AND column_name = 'downloadSource'
+    ) THEN
+        ALTER TABLE "SystemSettings" ADD COLUMN "downloadSource" TEXT NOT NULL DEFAULT 'soulseek';
+    END IF;
+END $$;
+
+-- Add primaryFailureFallback column if it doesn't exist (idempotent)
+DO $$
+BEGIN
+    IF NOT EXISTS (
+        SELECT 1 FROM information_schema.columns
+        WHERE table_name = 'SystemSettings' AND column_name = 'primaryFailureFallback'
+    ) THEN
+        ALTER TABLE "SystemSettings" ADD COLUMN "primaryFailureFallback" TEXT NOT NULL DEFAULT 'none';
+    END IF;
+END $$;

backend/src/routes/artists.ts

@@ -5,6 +5,7 @@ import { musicBrainzService } from "../services/musicbrainz";
 import { fanartService } from "../services/fanart";
 import { deezerService } from "../services/deezer";
 import { redisClient } from "../utils/redis";
+import { normalizeToArray } from "../utils/normalize";
 
 const router = Router();
 
@@ -158,8 +159,10 @@ router.get("/discover/:nameOrMbid", async (req, res) => {
         }
 
         // Fallback to Last.fm (but filter placeholders)
+        // NORMALIZATION: lastFmInfo.image could be a single object or array
         if (!image && lastFmInfo?.image) {
-            const lastFmImage = lastFmService.getBestImage(lastFmInfo.image);
+            const images = normalizeToArray(lastFmInfo.image);
+            const lastFmImage = lastFmService.getBestImage(images);
             // Filter out Last.fm placeholder
             if (
                 lastFmImage &&
@@ -274,10 +277,13 @@ router.get("/discover/:nameOrMbid", async (req, res) => {
         }
 
         // Get similar artists from Last.fm and fetch images
-        const similarArtistsRaw = lastFmInfo?.similar?.artist || [];
+        // NORMALIZATION: lastFmInfo.similar.artist could be a single object or array
+        const similarArtistsRaw = normalizeToArray(lastFmInfo?.similar?.artist);
         const similarArtists = await Promise.all(
             similarArtistsRaw.slice(0, 10).map(async (artist: any) => {
-                const similarImage = artist.image?.find(
+                // NORMALIZATION: artist.image could be a single object or array
+                const images = normalizeToArray(artist.image);
+                const similarImage = images.find(
                     (img: any) => img.size === "large"
                 )?.[" #text"];
 
@@ -325,14 +331,19 @@ router.get("/discover/:nameOrMbid", async (req, res) => {
             })
         );
 
+        // NORMALIZATION: lastFmInfo.tags.tag could be a single object or array
+        const tags = normalizeToArray(lastFmInfo?.tags?.tag)
+            .map((t: any) => t?.name)
+            .filter(Boolean);
+
         const response = {
             mbid,
             name: artistName,
             image,
             bio, // Use filtered bio instead of raw Last.fm bio
             summary: bio, // Alias for consistency
-            tags: lastFmInfo?.tags?.tag?.map((t: any) => t.name) || [],
+            tags,
-            genres: lastFmInfo?.tags?.tag?.map((t: any) => t.name) || [], // Alias for consistency
+            genres: tags, // Alias for consistency
             listeners: parseInt(lastFmInfo?.stats?.listeners || "0"),
             playcount: parseInt(lastFmInfo?.stats?.playcount || "0"),
             url: lastFmInfo?.url || null,
@@ -470,7 +481,10 @@ router.get("/album/:mbid", async (req, res) => {
 
         // Check if Cover Art Archive actually has the image
         try {
-            const response = await fetch(coverArtUrl, { method: "HEAD" });
+            const response = await fetch(coverArtUrl, {
+                method: "HEAD",
+                signal: AbortSignal.timeout(2000),
+            });
             if (response.ok) {
                 coverUrl = coverArtUrl;
                 logger.debug(`Cover Art Archive has cover for ${albumTitle}`);
@@ -529,7 +543,10 @@ router.get("/album/:mbid", async (req, res) => {
             coverUrl,
             coverArt: coverUrl, // Alias for compatibility
             bio: lastFmInfo?.wiki?.summary || null,
-            tags: lastFmInfo?.tags?.tag?.map((t: any) => t.name) || [],
+            // NORMALIZATION: lastFmInfo.tags.tag could be a single object or array
+            tags: normalizeToArray(lastFmInfo?.tags?.tag)
+                .map((t: any) => t?.name)
+                .filter(Boolean),
             tracks: tracks.map((track: any, index: number) => ({
                 id: `mb-${releaseGroupId}-${track.id || index}`,
                 title: track.title,

backend/src/routes/library.ts

@@ -2754,32 +2754,12 @@ router.get("/tracks/:id/stream", async (req, res) => {
                     `[STREAM] Sending file: ${filePath}, mimeType: ${mimeType}`
                 );
 
-                res.sendFile(
+                await streamingService.streamFileWithRangeSupport(req, res, filePath, mimeType);
-                    filePath,
+                streamingService.destroy();
-                    {
+                logger.debug(
-                        headers: {
+                    `[STREAM] File sent successfully: ${path.basename(
-                            "Content-Type": mimeType,
+                        filePath
-                            "Accept-Ranges": "bytes",
+                    )}`
-                            "Cache-Control": "public, max-age=31536000",
-                            "Access-Control-Allow-Origin":
-                                req.headers.origin || "*",
-                            "Access-Control-Allow-Credentials": "true",
-                            "Cross-Origin-Resource-Policy": "cross-origin",
-                        },
-                    },
-                    (err) => {
-                        // Always destroy the streaming service to clean up intervals
-                        streamingService.destroy();
-                        if (err) {
-                            logger.error(`[STREAM] sendFile error:`, err);
-                        } else {
-                            logger.debug(
-                                `[STREAM] File sent successfully: ${path.basename(
-                                    filePath
-                                )}`
-                            );
-                        }
-                    }
                 );
 
                 return;
@@ -2812,30 +2792,8 @@ router.get("/tracks/:id/stream", async (req, res) => {
                             absolutePath
                         );
 
-                    res.sendFile(
+                    await streamingService.streamFileWithRangeSupport(req, res, filePath, mimeType);
-                        filePath,
+                    streamingService.destroy();
-                        {
-                            headers: {
-                                "Content-Type": mimeType,
-                                "Accept-Ranges": "bytes",
-                                "Cache-Control": "public, max-age=31536000",
-                                "Access-Control-Allow-Origin":
-                                    req.headers.origin || "*",
-                                "Access-Control-Allow-Credentials": "true",
-                                "Cross-Origin-Resource-Policy": "cross-origin",
-                            },
-                        },
-                        (err) => {
-                            // Always destroy the streaming service to clean up intervals
-                            streamingService.destroy();
-                            if (err) {
-                                logger.error(
-                                    `[STREAM] sendFile fallback error:`,
-                                    err
-                                );
-                            }
-                        }
-                    );
                     return;
                 }
 

backend/src/services/audioStreaming.ts

@@ -1,4 +1,6 @@
 import * as fs from "fs";
+import { promises as fsPromises } from "fs";
+import { Request, Response } from "express";
 import { logger } from "../utils/logger";
 import * as path from "path";
 import * as crypto from "crypto";
@@ -384,6 +386,95 @@ export class AudioStreamingService {
         return mimeTypes[ext] || "audio/mpeg";
     }
 
+    /**
+     * Stream file with proper HTTP Range support (fixes Firefox FLAC issue #42/#17)
+     * Manually handles Range requests to ensure compatibility with Firefox's strict
+     * Content-Range header validation for large FLAC files.
+     */
+    async streamFileWithRangeSupport(
+        req: Request,
+        res: Response,
+        filePath: string,
+        mimeType: string
+    ): Promise<void> {
+        try {
+            // Get file stats for size
+            const stats = await fsPromises.stat(filePath);
+            const fileSize = stats.size;
+
+            // Parse Range header
+            const range = req.headers.range;
+            let start = 0;
+            let end = fileSize - 1;
+
+            if (range) {
+                // Parse bytes=START-END or bytes=START-
+                const parts = range.replace(/bytes=/, "").split("-");
+                start = parseInt(parts[0], 10);
+                end = parts[1] ? parseInt(parts[1], 10) : fileSize - 1;
+
+                // Validate range
+                if (start >= fileSize || end >= fileSize || start > end) {
+                    res.status(416).set({
+                        "Content-Range": `bytes */${fileSize}`,
+                    });
+                    res.end();
+                    return;
+                }
+            }
+
+            const contentLength = end - start + 1;
+
+            // Set response headers
+            const headers: Record<string, string> = {
+                "Content-Type": mimeType,
+                "Accept-Ranges": "bytes",
+                "Cache-Control": "public, max-age=31536000",
+                "Content-Length": contentLength.toString(),
+            };
+
+            // Add CORS headers from request origin
+            if (req.headers.origin) {
+                headers["Access-Control-Allow-Origin"] = req.headers.origin;
+                headers["Access-Control-Allow-Credentials"] = "true";
+            }
+
+            // Set status and range-specific headers
+            if (range) {
+                res.status(206);
+                headers["Content-Range"] = `bytes ${start}-${end}/${fileSize}`;
+            } else {
+                res.status(200);
+            }
+
+            res.set(headers);
+
+            // Create read stream with range
+            const stream = fs.createReadStream(filePath, { start, end });
+
+            // Handle stream errors
+            stream.on("error", (err) => {
+                logger.error(`[AudioStreaming] Stream error for ${filePath}:`, err);
+                if (!res.headersSent) {
+                    res.status(500).end();
+                }
+            });
+
+            // Handle cleanup on response close
+            res.on("close", () => {
+                stream.destroy();
+            });
+
+            // Pipe stream to response
+            stream.pipe(res);
+        } catch (err) {
+            logger.error(`[AudioStreaming] Failed to stream ${filePath}:`, err);
+            if (!res.headersSent) {
+                res.status(500).end();
+            }
+        }
+    }
+
     /**
      * Cleanup resources
      */

backend/src/services/soulseek.ts

@@ -6,6 +6,7 @@
 import slsk from "slsk-client";
 import path from "path";
 import fs from "fs";
+import { mkdir } from "fs/promises";
 import PQueue from "p-queue";
 import { getSystemSettings } from "../utils/systemSettings";
 import { sessionLog } from "../utils/playlistLogger";
@@ -700,10 +701,14 @@ class SoulseekService {
             return { success: false, error: "Not connected" };
         }
 
-        // Ensure destination directory exists
+        // Ensure destination directory exists (idempotent - won't fail if exists)
         const destDir = path.dirname(destPath);
-        if (!fs.existsSync(destDir)) {
+        try {
-            fs.mkdirSync(destDir, { recursive: true });
+            await mkdir(destDir, { recursive: true });
+        } catch (err: any) {
+            sessionLog("SOULSEEK", `Failed to create directory ${destDir}: ${err.message}`, "ERROR");
+            this.activeDownloads--;
+            return { success: false, error: `Cannot create destination directory: ${err.message}` };
         }
 
         sessionLog(

backend/src/services/wikidata.ts

@@ -78,14 +78,11 @@ class WikidataService {
       LIMIT 1
     `;
 
-        const response = await axios.get("https://query.wikidata.org/sparql", {
+        const response = await this.client.get("https://query.wikidata.org/sparql", {
             params: {
                 query: sparqlQuery,
                 format: "json",
             },
-            headers: {
-                "User-Agent": "Lidify/1.0.0",
-            },
         });
 
         const bindings = response.data.results?.bindings || [];
@@ -100,7 +97,7 @@ class WikidataService {
     ): Promise<string | undefined> {
         try {
             // Get English Wikipedia article title
-            const response = await axios.get(
+            const response = await this.client.get(
                 `https://www.wikidata.org/wiki/Special:EntityData/${wikidataId}.json`
             );
 
@@ -110,7 +107,7 @@ class WikidataService {
             if (!enWikiTitle) return undefined;
 
             // Get article summary from Wikipedia API
-            const summaryResponse = await axios.get(
+            const summaryResponse = await this.client.get(
                 "https://en.wikipedia.org/api/rest_v1/page/summary/" +
                     encodeURIComponent(enWikiTitle)
             );
@@ -129,7 +126,7 @@ class WikidataService {
         wikidataId: string
     ): Promise<string | undefined> {
         try {
-            const response = await axios.get(
+            const response = await this.client.get(
                 `https://www.wikidata.org/wiki/Special:EntityData/${wikidataId}.json`
             );
 

backend/src/workers/organizeSingles.ts

@@ -121,9 +121,12 @@ async function migrateExistingSoulseekFiles(musicPath: string): Promise<void> {
                 continue;
             }
 
-            // Create destination directory
+            // Create destination directory (idempotent - won't fail if exists)
-            if (!fs.existsSync(destDir)) {
+            try {
                 fs.mkdirSync(destDir, { recursive: true });
+            } catch (err: any) {
+                sessionLog('ORGANIZE', `Failed to create directory ${destDir}: ${err.message}`, 'WARN');
+                continue; // Skip this file, try next
             }
 
             // Move file (copy then delete original)

frontend/components/layout/AuthenticatedLayout.tsx

@@ -129,7 +129,7 @@ export function AuthenticatedLayout({ children }: { children: ReactNode }) {
                                 tabIndex={-1}
                                 className="flex-1 bg-gradient-to-b from-[#1a1a1a] via-black to-black mx-2 mb-2 rounded-lg overflow-y-auto relative focus:outline-none"
                                 style={{
-                                    marginTop: "58px",
+                                    marginTop: "calc(58px + env(safe-area-inset-top, 0px))",
                                     marginBottom:
                                         "calc(56px + env(safe-area-inset-bottom, 0px) + 8px)",
                                 }}

frontend/components/layout/Sidebar.tsx

@@ -17,6 +17,7 @@ const navigation = [
     { name: "Library", href: "/library" },
     { name: "Radio", href: "/radio" },
     { name: "Discovery", href: "/discover" },
+    { name: "Releases", href: "/releases" },
     { name: "Audiobooks", href: "/audiobooks" },
     { name: "Podcasts", href: "/podcasts" },
     { name: "Browse", href: "/browse/playlists", badge: "Beta" },

frontend/components/layout/TopBar.tsx

@@ -142,7 +142,10 @@ export function TopBar() {
     return (
         <header
             className="fixed top-0 left-0 right-0 bg-black flex items-center px-3 z-50"
-            style={{ height: isMobileOrTablet ? "58px" : "64px" }}
+            style={{
+                height: isMobileOrTablet ? "58px" : "64px",
+                paddingTop: isMobileOrTablet ? "env(safe-area-inset-top)" : undefined,
+            }}
         >
             {/* Mobile/Tablet Layout: Hamburger + Home + Search + Bell */}
             {isMobileOrTablet ? (

frontend/components/player/SeekSlider.tsx

@@ -247,7 +247,8 @@ export function SeekSlider({
         >
             <div
                 className={cn(
-                    "h-full rounded-full relative transition-all duration-150",
+                    "h-full rounded-full relative",
+                    !isDragging && "transition-all duration-150",
                     styles.progress
                 )}
                 style={{ width: `${displayProgress}%` }}

frontend/components/ui/PullToRefresh.tsx

@@ -13,6 +13,11 @@ export function PullToRefresh({
     children,
     threshold = 80,
 }: PullToRefreshProps) {
+    // HOTFIX v1.3.2: Temporarily disabled - blocking mobile scrolling
+    // TODO: Fix in v1.4 - Issues: 1) h-full breaks flex layout, 2) touch handlers may interfere
+    // Proper fix: Change line 90 className to "relative flex-1 flex flex-col min-h-0"
+    return <>{children}</>;
+
     const [pullDistance, setPullDistance] = useState(0);
     const [isRefreshing, setIsRefreshing] = useState(false);
     const startY = useRef(0);

frontend/lib/api.ts

@@ -1,4 +1,5 @@
 const AUTH_TOKEN_KEY = "auth_token";
+const REFRESH_TOKEN_KEY = "refresh_token";
 
 // Mood Mix Types (Legacy - for old presets endpoint)
 export interface MoodPreset {
@@ -113,6 +114,7 @@ class ApiClient {
             if (this.token) {
                 this.tokenInitialized = true;
             }
+            // Note: Refresh token is loaded on-demand via getRefreshToken()
         }
     }
 
@@ -153,19 +155,31 @@ class ApiClient {
         this.baseUrl = "";
     }
 
-    // Store JWT token
+    // Store JWT token and optionally refresh token
-    setToken(token: string) {
+    setToken(token: string, refreshToken?: string) {
         this.token = token;
         if (typeof window !== "undefined") {
             localStorage.setItem(AUTH_TOKEN_KEY, token);
+            if (refreshToken) {
+                localStorage.setItem(REFRESH_TOKEN_KEY, refreshToken);
+            }
         }
     }
 
-    // Clear JWT token
+    // Get refresh token from storage
+    getRefreshToken(): string | null {
+        if (typeof window === "undefined") {
+            return null;
+        }
+        return localStorage.getItem(REFRESH_TOKEN_KEY);
+    }
+
+    // Clear both JWT tokens
     clearToken() {
         this.token = null;
         if (typeof window !== "undefined") {
             localStorage.removeItem(AUTH_TOKEN_KEY);
+            localStorage.removeItem(REFRESH_TOKEN_KEY);
         }
     }
 
@@ -177,15 +191,56 @@ class ApiClient {
         return getApiBaseUrl();
     }
 
+    /**
+     * Refresh the access token using the refresh token
+     * @returns true if refresh succeeded, false otherwise
+     */
+    private async refreshAccessToken(): Promise<boolean> {
+        const refreshToken = this.getRefreshToken();
+        if (!refreshToken) {
+            return false;
+        }
+
+        try {
+            const response = await fetch(`${this.getBaseUrl()}/api/auth/refresh`, {
+                method: "POST",
+                headers: { "Content-Type": "application/json" },
+                body: JSON.stringify({ refreshToken }),
+                credentials: "include",
+            });
+
+            if (!response.ok) {
+                // Refresh token invalid or expired - clear tokens
+                this.clearToken();
+                return false;
+            }
+
+            const data = await response.json();
+            
+            // Store new tokens
+            if (data.token) {
+                this.setToken(data.token, data.refreshToken);
+                return true;
+            }
+            
+            this.clearToken();
+            return false;
+        } catch (error) {
+            console.error("[API] Token refresh failed:", error);
+            this.clearToken();
+            return false;
+        }
+    }
+
     /**
      * Make an authenticated API request
      * Public method for components that need custom API calls
      */
     async request<T>(
         endpoint: string,
-        options: RequestInit & { silent404?: boolean } = {}
+        options: RequestInit & { silent404?: boolean; _retryCount?: number } = {}
     ): Promise<T> {
-        const { silent404, ...fetchOptions } = options;
+        const { silent404, _retryCount = 0, ...fetchOptions } = options;
         const headers: HeadersInit = {
             "Content-Type": "application/json",
             ...fetchOptions.headers,
@@ -217,6 +272,23 @@ class ApiClient {
                 console.error(`[API] Request failed: ${url}`, error);
             }
 
+            // Handle 401 with token refresh (retry once)
+            if (response.status === 401 && _retryCount === 0 && endpoint !== "/auth/refresh") {
+                console.log("[API] 401 error - attempting token refresh");
+                const refreshed = await this.refreshAccessToken();
+                
+                if (refreshed) {
+                    console.log("[API] Token refreshed - retrying request");
+                    // Retry the request with new token
+                    return this.request<T>(endpoint, {
+                        ...options,
+                        _retryCount: 1, // Prevent infinite loops
+                    });
+                }
+                
+                console.log("[API] Token refresh failed - user needs to re-login");
+            }
+
             if (response.status === 401) {
                 const err = new Error("Not authenticated");
                 (err as any).status = response.status;
@@ -260,6 +332,7 @@ class ApiClient {
     async login(username: string, password: string, token?: string) {
         const data = await this.request<{
             token?: string;
+            refreshToken?: string;
             user?: {
                 id: string;
                 username: string;
@@ -274,9 +347,9 @@ class ApiClient {
             body: JSON.stringify({ username, password, token }),
         });
 
-        // If login returned a JWT token, store it
+        // If login returned JWT tokens, store them
         if (data.token) {
-            this.setToken(data.token);
+            this.setToken(data.token, data.refreshToken);
         }
 
         // Return user data in consistent format

services/audio-analyzer/analyzer.py

@@ -65,6 +65,7 @@ import traceback
 import numpy as np
 from concurrent.futures import ProcessPoolExecutor, as_completed
 import multiprocessing
+import gc
 
 # BrokenProcessPool was added in Python 3.9, provide compatibility for Python 3.8
 try:
@@ -112,6 +113,14 @@ except ImportError as e:
 TF_MODELS_AVAILABLE = False
 TensorflowPredictMusiCNN = None
 try:
+    import tensorflow as tf
+    # Limit TensorFlow memory usage (CPU & GPU)
+    try:
+        gpus = tf.config.experimental.list_physical_devices('GPU')
+        for gpu in gpus:
+            tf.config.experimental.set_memory_growth(gpu, True)
+    except Exception:
+        pass
     from essentia.standard import TensorflowPredictMusiCNN
     TF_MODELS_AVAILABLE = True
     logger.info("TensorflowPredictMusiCNN available - Enhanced mode enabled")
@@ -376,14 +385,18 @@ class AudioAnalyzer:
             traceback.print_exc()
             self.enhanced_mode = False
     
-    def load_audio(self, file_path: str, sample_rate: int = 16000) -> Optional[Any]:
+    def load_audio(self, file_path: str, sample_rate: int = 16000, max_duration: int = 90) -> Optional[Any]:
-        """Load audio file as mono signal"""
+        """Load up to max_duration seconds of audio as mono signal (to limit memory usage)"""
         if not ESSENTIA_AVAILABLE:
             return None
         
         try:
             loader = es.MonoLoader(filename=file_path, sampleRate=sample_rate)
             audio = loader()
+            # Limit to max_duration seconds
+            max_samples = int(sample_rate * max_duration)
+            if len(audio) > max_samples:
+                audio = audio[:max_samples]
             return audio
         except Exception as e:
             logger.error(f"Failed to load audio {file_path}: {e}")
@@ -514,12 +527,17 @@ class AudioAnalyzer:
             result['_error'] = 'Essentia library not installed'
             return result
         
-        # Load audio at different sample rates for different algorithms
+        # Limit memory: only analyze up to MAX_ANALYZE_SECONDS (default 90s)
-        audio_44k = self.load_audio(file_path, 44100)
+        MAX_ANALYZE_SECONDS = int(os.getenv('MAX_ANALYZE_SECONDS', '90'))
-        audio_16k = self.load_audio(file_path, 16000)
+        try:
-        
+            # Load audio at different sample rates for different algorithms, limit duration
+            audio_44k = self.load_audio(file_path, 44100, max_duration=MAX_ANALYZE_SECONDS)
+            audio_16k = self.load_audio(file_path, 16000, max_duration=MAX_ANALYZE_SECONDS)
+        except MemoryError:
+            logger.error(f"MemoryError: Could not load audio for {file_path}")
+            result['_error'] = 'MemoryError: audio file too large'
+            return result
         if audio_44k is None or audio_16k is None:
-            result['_error'] = 'Failed to load audio file'
             return result
         
         # Validate audio before analysis (Phase 2 defensive improvement)
@@ -586,7 +604,10 @@ class AudioAnalyzer:
             # Process audio in frames for detailed analysis
             frame_size = 2048
             hop_size = 1024
-            for i in range(0, len(audio_44k) - frame_size, hop_size):
+            max_frames = int((44100 * MAX_ANALYZE_SECONDS - frame_size) / hop_size)
+            for idx, i in enumerate(range(0, len(audio_44k) - frame_size, hop_size)):
+                if idx > max_frames:
+                    break
                 frame = audio_44k[i:i + frame_size]
                 windowed = self.windowing(frame)
                 spectrum = self.spectrum(windowed)
@@ -599,7 +620,6 @@ class AudioAnalyzer:
             # RMS-based energy (properly normalized to 0-1)
             if rms_values:
                 avg_rms = np.mean(rms_values)
-                # RMS is typically 0.0-0.5 for normalized audio, scale to 0-1
                 result['energy'] = round(min(1.0, float(avg_rms) * 3), 3)
             else:
                 result['energy'] = 0.5
@@ -616,7 +636,6 @@ class AudioAnalyzer:
             result['_zcr'] = np.mean(zcr_values) if zcr_values else 0.1
             
             # Basic Danceability (non-ML)
-            # Note: es.Danceability() can return values > 1.0, so we clamp
             danceability, _ = self.danceability_extractor(audio_44k)
             result['danceability'] = round(max(0.0, min(1.0, float(danceability))), 3)
             
@@ -632,22 +651,25 @@ class AudioAnalyzer:
                     traceback.print_exc()
                     self._apply_standard_estimates(result, scale, bpm)
             else:
-                # === STANDARD MODE: Use heuristics ===
                 self._apply_standard_estimates(result, scale, bpm)
             
             # Generate mood tags based on all features
             result['moodTags'] = self._generate_mood_tags(result)
             
             logger.info(f"Analysis complete [{result['analysisMode']}]: BPM={result['bpm']}, Key={result['key']} {result['keyScale']}, Valence={result['valence']}, Arousal={result['arousal']}")
-            
+        except MemoryError:
+            logger.error(f"MemoryError during analysis of {file_path}")
+            result['_error'] = 'MemoryError: analysis exceeded memory limits'
         except Exception as e:
             logger.error(f"Analysis error: {e}")
             traceback.print_exc()
-        
+        finally:
-        # Clean up internal fields before returning
+            # Clean up internal fields before returning
-        for key in ['_spectral_centroid', '_spectral_flatness', '_zcr']:
+            for key in ['_spectral_centroid', '_spectral_flatness', '_zcr']:
-            result.pop(key, None)
+                result.pop(key, None)
-        
+            # Explicitly free memory
+            del audio_44k, audio_16k
+            gc.collect()
         return result
     
     def _extract_ml_features(self, audio_16k) -> Dict[str, Any]: