Skip to content

v0.5.0

Choose a tag to compare

View all tags
@simongonzalezdc simongonzalezdc released this 23 Mar 01:38
· 501 commits to master since this release
v0.5.0
b4a9c38

mcp-video v0.5.0 Release Notes

🎉 What's New

11 new FFmpeg features across 5 development waves, plus comprehensive testing and bug fixes.

New MCP Tools (11)

Wave 1: Enhanced Editing

  • Ken Burns effect - Smooth pan/zoom animation filter for dynamic video movement
  • Two-pass encoding - Higher quality at lower bitrates for efficient compression

Wave 2: Audio Effects

  • Audio reverb - Add echo/reverb effect to audio tracks
  • Audio compressor - Dynamic range compression for balanced audio
  • Pitch shift - Change audio pitch while maintaining duration
  • Noise reduction - Remove background noise from audio tracks

Wave 3: Analysis & Export

  • Scene detection - Automatically identify scene changes in videos
  • Image sequences - Create videos from image sequences, export frames
  • Quality metrics - PSNR/SSIM comparison between original and processed videos
  • Metadata editing - Read and write video metadata tags

Wave 4: Advanced Processing

  • Video stabilization - Stabilize shaky footage using motion vectors (requires vidstab)
  • Apply mask - Image masking with edge feathering for compositing

Wave 5: Content Generation

  • Generate subtitles - Create SRT subtitles from text entries, optionally burn into video
  • Audio waveform - Extract audio waveform data (peaks and silence regions)

API Improvements

Client Class

  • New methods: detect_scenes(), create_from_images(), export_frames(), generate_subtitles(), audio_waveform(), compare_quality(), read_metadata(), write_metadata(), stabilize(), apply_mask(), batch()
  • convert() now supports two_pass and target_bitrate parameters

CLI Enhancements

  • Rich terminal output with tables, spinners, and styled error panels
  • --format json mode for scripted/pipe-friendly JSON output
  • Video templates: TikTok, YouTube Shorts, Instagram Reel, YouTube, Instagram Post
  • New commands: detect-scenes, create-from-images, export-frames, generate-subtitles, audio-waveform, compare-quality, read-metadata, write-metadata, stabilize, apply-mask, batch, templates

Bug Fixes

  • Fixed FFmpeg return code checking in detect_scenes() and compare_quality()
  • Fixed metadata validation to check both keys and values
  • Fixed audio_waveform() to check FFmpeg return code before parsing
  • Fixed vidstabtransform parameter: zooming → zoom
  • Fixed create_from_images() to normalize images before concatenation

Testing

545 tests passing, including:

  • 38 engine tests
  • 119 server tests
  • 413 adversarial/red-team tests
  • 331 CLI tests
  • 18 comprehensive real-video integration tests

Installation

pip install --upgrade mcp-video

Or with UVX:

uvx mcp_video

Documentation

  • Updated README with all 29 tools
  • Added CLI reference
  • Added template documentation
  • Roadmap updated with completed features

Breaking Changes

None. Fully backward compatible with v0.4.x.

Requirements

  • Python 3.11+
  • FFmpeg (for video processing)
  • Optional: FFmpeg with vidstab (for stabilization feature)

Full Changelog

Added

  • 11 new MCP tools for advanced video processing
  • 5 result types: SceneDetectionResult, ImageSequenceResult, SubtitleResult, WaveformResult, QualityMetricsResult, MetadataResult
  • Rich CLI output with formatting options
  • Video templates for social media platforms
  • Comprehensive real-video testing suite

Changed

  • Client.convert() now supports two_pass and target_bitrate
  • Improved error messages with auto-fix suggestions
  • Better validation for metadata operations

Fixed

  • FFmpeg return code checking in detect_scenes and compare_quality
  • Metadata value validation for special characters
  • audio_waveform return code checking
  • vidstabtransform parameter name (zooming → zoom)
  • create_from_images image normalization

Contributors

  • @pastorsimon1798

Links

Contributors

simongonzalezdc
Assets 2
Loading