CapCut Desktop Automation & Timeline Director Skill

This skill equips AI assistants (such as Claude Desktop, Cursor, Google Antigravity, Cline, and Windsurf) with the operational knowledge required to edit videos using CapCut Desktop or JianYing Pro through the CapCut MCP Server.

Core Capabilities

1. Project & Lifecycle Management:
   • Discover local projects (capcut_list_projects).
   • Create fresh drafts with custom canvas resolutions (capcut_create_project).
   • Duplicate or delete projects (capcut_duplicate_project, capcut_delete_project).
   • Launch projects directly in CapCut Desktop (capcut_open_app).
2. Multi-Track Timeline Construction:
   • Insert video clips, B-roll cutaways, and talking-head footage (capcut_add_video_clip).
   • Insert background music, sound effects, and voiceovers (capcut_add_audio_clip, capcut_add_sound_effect).
   • Trim, split, and reorder clips (capcut_trim_clip, capcut_split_clip, capcut_remove_clip).
3. Motion Design & Transformations:
   • 2D transforms: Position (x, y), Scale (1.0 = 100%), Rotation (degrees) (capcut_set_clip_transform).
   • Dynamic Speed Ramping & Easing: Set speed multiplier (0.1x to 100.0x) with optional tone-correction (capcut_set_clip_speed).
   • Keyframe Animation: Animate scale, opacity, and coordinates over time (capcut_add_keyframe).
   • Seamless Transitions: Add dissolves, push, zoom, and whip transitions between clips (capcut_add_transition).
4. Viral Typography & Subtitle Engine:
   • Render synchronized SRT subtitles directly to native text tracks (capcut_add_subtitles).
   • Generate viral short-form typography with custom colors, strokes, and drop shadows (capcut_add_viral_caption).
   • Standard static and lower-third text elements (capcut_add_text).
5. Headless Verification:
   • Inspect media resolution, duration, and audio channels (capcut_get_media_info).
   • Render fast proxy preview MP4s headlessly without launching the GUI (capcut_generate_preview).

Timeline Coordinate System & Invariants

When manipulating CapCut projects, the following invariants apply:

1. Time Units: Microseconds

• All timestamps (start_time, duration, source_start, source_duration) are measured in microseconds ($1\text{s} = 1,000,000,\mu\text{s}$).
• Example: A 3.5-second clip has a duration of 3500000.

2. Canvas Coordinates

• Canvas center is (0.0, 0.0).
• Normalized range:
  • Horizontal (x): [-1.0, 1.0] (left to right).
  • Vertical (y): [-1.0, 1.0] (bottom to top, where 0.0 is middle, 0.6 is top third, -0.6 is bottom third).
• Standard scale: 1.0 = 100% canvas fit.

3. Track Hierarchy

• Track 0 (Base Video): Main A-roll / talking-head continuous footage.
• Track 1+ (Overlay Video): B-roll cutaways, screen recordings, PIP (picture-in-picture).
• Text Track 0: Primary captions or synchronized subtitles.
• Text Track 1+: Floating titles, lower-thirds, emoji stickers.
• Audio Track 0: Background music (ducked to -12dB).
• Audio Track 1+: Sound design (whooshes, pops, impacts, risers).

Viral Typography Styles Reference

When styling short-form captions with capcut_add_viral_caption:

Style 6: Glow Pop

• Font: Bold Grotesk / Montserrat Black / Impact.
• Primary Color: High-saturation electric yellow (#FFE600) or lime (#00FF66).
• Stroke: Heavy black outline (#000000, width 12-16).
• Shadow: Deep black drop shadow with soft blur.
• Motion: Scale pop (1.15x -> 1.0x) over 120ms at word boundary.

Style 7: Hero Typography

• Letter Spacing: -1.0 (tight letter tracking).
• Casing: Uppercase.
• Shadow Offset: (26, 26) with 85% opacity.
• Color Accent: Single focal word highlighted in cyan (#00E5FF) or electric magenta (#FF0055).

Autonomous Editing Workflow Playbook

When an AI agent is requested to edit a raw video into a complete video reel, follow this exact sequence:

1. Inspect Media: Call capcut_get_media_info on the user's raw files to obtain precise duration, width, height, and FPS.
2. Create Project: Call capcut_create_project with name and target canvas resolution (e.g., 1080x1920 for 9:16 vertical shorts, or 1920x1080 for 16:9 widescreen).
3. Assemble A-Roll: Call capcut_add_video_clip to place the primary speech track onto track index 0.
4. Add Subtitles / Captions: Call capcut_add_subtitles with an SRT payload, or use capcut_add_viral_caption for styled kinetic text callouts.
5. Insert B-Roll Overlays: Call capcut_add_video_clip with track_index: 1 over key narrative beats. Apply punch-in zoom transforms (scale: 1.12) where appropriate.
6. Add Background Music & Sound Effects: Call capcut_add_audio_clip with volume: 0.25 (-12dB ducking). Call capcut_add_sound_effect at cutaway points (e.g. whoosh on B-roll entrance).
7. Verify or Open: Generate a quick proxy review using capcut_generate_preview, or open the project directly in the desktop app for the user using capcut_open_app.

Reference Manuals

For deep video engineering standards, mathematical Quad Easing formulas, and Walter Murch's Rule of Six, consult:

• CapCut Master Manual
• B-Roll Production Manual