Skip to content

CHANGELOG.md

Latest commit

 

History

History
135 lines (116 loc) · 5.96 KB

CHANGELOG.md

File metadata and controls

135 lines (116 loc) · 5.96 KB
Raw

Changelog

All notable changes to this project will be documented in this file.

The format is based on Keep a Changelog, and this project adheres to Semantic Versioning.

[1.5.1] - 2025-10-17

Fixed

  • GitHub Release Automation - Fixed automated release process to enable installation scripts
    • Ensured GitHub Actions workflow runs on tag push
    • Creates proper GitHub Release with all platform binaries
    • Fixes installation scripts that depend on GitHub Releases

Documentation

  • User Documentation Suite - Comprehensive documentation for end users
    • Created INSTALL.md - Platform-specific installation guide
    • Created QUICKSTART.md - 5-minute getting started guide
    • Created MCP_SETUP.md - Claude Code/Desktop integration setup
    • Created TROUBLESHOOTING.md - Common issues reference
    • Updated README.md to user-friendly landing page
    • Created RELEASE.md - Release process documentation

Maintenance

  • Repository Cleanup - Removed legacy files and improved organization
    • Removed lfm-mcp-prototype directory (2,061 files including node_modules)
    • Removed old planning docs and test files from git
    • Moved development docs to docs/ folder
    • Updated .gitignore to prevent re-adding legacy files

1.5.0 - 2025-10-17

Added

MCP Server Enhancements (v0.2.0)

  • "Listen Before You Speak" methodology - Core workflow guidance for LLM interactions
    • Emphasizes data → think → more data → think → narrative pattern
    • Prevents starting narratives before completing research phase
  • "Using lfm_recommendations - LLM Reasoning First" - Guidelines for proper tool usage
    • Emphasizes LLM musical knowledge over algorithmic recommendations
    • Clear guidance on when NOT to use recommendations (well-known artists)
    • Proper workflow: LLM reasoning first, optionally supplement with tool
  • "Playback State Awareness" - Mandatory workflow before playback suggestions
    • Trigger words identification ("I've queued", "Want me to queue", etc.)
    • Required state checks: recent_tracks + current_track before suggesting playback
    • Response templates for data-informed suggestions

Features

  • Sonos Integration - Full playback support via node-sonos-http-api
    • Unified play command supporting both Spotify and Sonos
    • Room management and device routing
    • Config-driven defaults (DefaultPlayer, DefaultRoom)
  • Recent Tracks Feature - Temporal listening history with chronological ordering
    • Most recent first ordering
    • Configurable hours lookback parameter
    • Perfect for detecting listening patterns
  • Spotify Playback Controls - Comprehensive playback management
    • Play now / queue functionality
    • Pause, resume, skip commands
    • Current track status
    • Device activation support
  • Album Disambiguation - Handle tracks on multiple albums
    • Detect and report multiple album versions
    • Allow track + album parameter combination
    • User preference for studio vs live/greatest hits versions
  • API Status Checker - Diagnose Last.fm connectivity issues
    • Health status for multiple endpoints
    • Verbose mode with HTTP details
    • JSON output for monitoring/automation
  • Parallel API Processing - Batch processing for deep searches
    • Configurable concurrent API calls (default: 5)
    • Intelligent rate limiting (1 second between batches)
    • Significant speedup for artist-tracks/artist-albums deep searches

Improved

  • MCP Parser - Fixed position-based JSON extraction for array-root responses
    • Correctly handles both object-root and array-root JSON structures
    • Helper functions: extractArray() and extractObject() with fallback logic
  • Token Optimization - ~50% reduction in MCP response tokens
    • Added compact helper functions (compactAlbum, compactArtist, compactTrack)
    • Strip URLs and MBIDs that LLMs can't use
    • Applied to lfm_albums, lfm_artists, lfm_tracks handlers
  • Spotify Authentication - Improved resilience
    • Changed StartPlaybackAsync signature from string to List
    • Ensures atomic album playback (all tracks in one API call)
    • Prevents race conditions in track ordering
  • Documentation Refactoring - Split CLAUDE.md into focused files
    • Created SESSION_HISTORY.md - 7 archived sessions
    • Created IMPLEMENTATION_NOTES.md - Technical reference
    • Created LESSONS_LEARNED.md - Debugging patterns and best practices
    • CLAUDE.md reduced by 53% while maintaining current work focus

Fixed

  • Album Queueing - Fixed missing/reordered tracks in Spotify album playback
    • Changed from 12 separate API calls to ONE atomic call
    • Supports up to 100 URIs in one request
  • Depth Parameter Clarification - Corrected understanding in guidelines
    • Depth = popularity ranking (top N by play count), NOT chronological
    • Updated guidelines to recommend deep:true as default
  • Cache Control - Added support for cache control in recent command
    • Respects cache expiry settings
    • Configurable throttling behavior

1.4.0 - Previous Release

Added

  • Enhanced cache system with comprehensive management
  • Comprehensive parameter validation and API improvements
  • Tag filtering system for recommendations
  • Recommendation diversity controls (totalArtists/totalTracks)
  • Artist-specific track and album search
  • Last.fm autocorrect support for artist name matching

Improved

  • Service layer architecture refactoring
  • Date range support across all commands
  • Unicode symbol support with auto-detection
  • API throttling configuration (default: 200ms)

1.0.0 - Initial Release

Added

  • Core Last.fm API integration
  • Basic commands: artists, tracks, albums
  • File-based caching system
  • Cross-platform support (Windows, Linux, WSL)
  • Configuration management
  • Spotify basic integration