| name | video-editing-basics |
| description | Use for video editing or video creation work using the Local Video Editor MCP server. Covers local video editing, trimming, captions, timeline operations, talking-head cleanup, B-roll, overlays, and export via local FFmpeg pipeline. |
Local Video Editor MCP โ Video Editing Assistant
Purpose
This skill is the base operating context for working with a local project through the local-video-editor MCP server. All processing runs entirely on your machine โ zero cloud dependency, file-based storage, local FFmpeg processing.
Use this for any video editing, trimming, captioning, or export task that should go through the Local Video Editor editing surface. Load the matching task-specific skill for detailed workflows.
Key Differences from Cloud-Based Editors
- Local stdio MCP server โ all tools are exposed via a local stdio MCP server, not a cloud API
- Zero cloud dependency โ no OAuth, no login, no cloud API keys required
- File-based storage โ media files are stored under
~/.local-video-editor/media/ and project data in ~/.local-video-editor/projects/
- FFmpeg-based pipeline โ export, frame extraction, and timeline composition use the system's local FFmpeg installation
- No upload required โ local files are referenced by filesystem path; no import to a remote bucket
Tool Naming Convention
All tools are prefixed with local-video-editor_. When referring to tools in instructions or workflows, always use the full local-video-editor_ prefix.
Examples:
local-video-editor_read_project instead of read_project
local-video-editor_edit_item instead of edit_item
local-video-editor_submit_export instead of submit_export
Available Tool Categories
The local-video-editor MCP provides tools in these categories:
| Category | Example Tools | Notes |
|---|
| Project management | create_project, list_projects, target_project, delete_project, get_editor_url | Full CRUD |
| Timeline & tracks | manage_timelines, read_timeline, edit_track, edit_item, split_item | Multi-track editing |
| Asset management | import_media, list_assets, get_asset | Local file imports |
| Script / A-roll | read_script, apply_script, clean_script | Transcript-based editing |
| Transcription | get_transcript, find_transcript, manage_transcript | Local Whisper ASR โ runs after import |
| Captions | edit_captions, read_captions | Enable, style, and template support |
| Frame inspection | view_timeline_frames, view_asset_frames | Local FFmpeg frame extraction |
| Export | submit_export, track_export, track_progress | Local FFmpeg rendering |
Role
When working in Local Video Editor projects, act as a professional video editing assistant. The user thinks in clips, cuts, stories, and visible outcomes, not data structures. Use video-editing judgment to clarify needs, recommend a concrete strategy, and execute the requested edit.
Limitations
The following capabilities are NOT available in Local Video Editor:
- ASR (speech-to-text) transcription โ transcription is available via local Whisper CLI. After importing media with audio, transcription runs automatically in the background.
manage_transcript supports fix for corrections and retry_transcription to re-run Whisper.
- AI generation tools โ no AI video generation, image generation, music generation, or voice synthesis
- Media/sound libraries โ
browse_library and search_stock_media are not available. Use local files only.
- Motion Graphic rendering โ MG code can be stored and edited but not rendered to video (Remotion-based rendering is not available)
- Cloud rendering โ all export is local FFmpeg.
export_motion_graphic_prores and request_asset_download are not available.
- OAuth / authentication โ no login, no API keys. The server runs as a local process.
Align Before Acting
Understanding the user's intended outcome is the foundation of creative editing. Clarify before committing to creative or strategic choices.
- "Make a 1-minute YouTube cut of this interview" โ may still need confirmation on what to keep
- "Cut this podcast into highlights" โ align on what counts as a highlight
- "Add captions" โ clear and narrow; execute
- "Make it shorter" after prior alignment โ usually does not need a new alignment round