Palmier Pro MCP Server

Lets an assistant edit video in Palmier Pro: timeline, clips, captions, color, audio and AI generation.

Official · Palmier Swift Media

68 popularity score ★ 15k

What it does

Palmier Pro is a macOS video editor with built-in AI generation, and while the app is open it runs an MCP server so agents such as Claude, Codex or Cursor can edit the open project on the timeline alongside the user.

Its tools read and inspect the timeline and media library, search media by what is on screen and what is said, import and organise media, place, insert, move, split, trim and remove clips, ripple-delete ranges, manage tracks, keyframes and layouts, sync clips and edit multicam groups, edit by transcript (remove words or silence), add and style text and captions, grade color and apply effects, denoise audio, detect beats, generate video, images and audio with AI models (which cost money), upscale media, export, and undo.

Download the app (macOS 26 on Apple Silicon only); the MCP server listens at http://127.0.0.1:19789/mcp over HTTP while the app is open. Claude Desktop can install a bundled desktop extension from the app's Help menu. Releases up to v0.7.6 and the published source are GPLv3; later binary releases are proprietary.

Tools 52

get_timeline
Always call at the start of a session.
inspect_timeline
See the composited timeline — what the user actually sees in the preview at a given frame: all video tracks stacked with their transforms, opacity, crop, edge softness, edge rounding, and keyframes applied, plus text and caption overlays baked in.
create_timeline
Creates a timeline and switches to it — every read and edit tool now targets it.
set_active_timeline
Switches the active timeline — the one every read and edit tool targets and the one the user sees.
manage_markers
Creates, updates, or deletes one persistent timeline marker.
set_project_settings
Change the project's frame rate, resolution, or aspect ratio.
export_project
Queues an export from the current project using the same modes as the Export dialog.
manage_exports
Lists or cancels exports for the current project.
get_media
The library inventory: media assets, folders, and timelines.
inspect_media
Look at a media asset before referencing or editing it.
search_media
Search the media library by content: what's on screen (visual) and what's said (spoken).
import_media
Imports external media into the project's library — the bridge for assets coming from other MCP servers (stock libraries, music services, web search) or local files the user already has.
capture_frame
Capture one video frame as a full-resolution PNG media asset.
organize_media
Reorganizes the library in one undoable action: create folders, move items into folders, rename items, delete items.
add_clips
Places one or more media assets on the timeline as a single undoable action.
insert_clips
Inserts one or more media assets at a single point and RIPPLES: every clip at or after atFrame is pushed right to open a gap, so nothing is overwritten.
move_clips
Moves one or more clips to a new track and/or frame position.
remove_clips
Removes one or more clips by ID as a single undoable action.
manage_clip_links
Links or unlinks clips as one undoable action without moving, trimming, or aligning them.
manage_tracks
Reorders, names, configures, or removes tracks in one undoable action.
split_clips
Splits clips into two at one or more cut points, all in a single undoable action.
ripple_delete_ranges
Cuts one or more ranges out and closes the gaps in one undoable action — the fast path for filler-word/dead-air removal.
swap_clip_media
Replace a clip's source with another library asset while preserving its edit, timing, framing, effects, and keyframes.
set_clip_properties
Apply the same generic clip property values to one or more clips in a single undoable action.
copy_clip_settings
Copy one clip's static settings to one or more clips of the same media type in a single undoable action.
set_keyframes
Set animated keyframes on one property of one clip.
apply_layout
Arrange multiple clips into a common multi-video layout (split screen, picture-in-picture, grid) in one undoable action — the fast path for composing several videos in one frame.
sync_clips
Align one or more clips to a reference clip by shifting targets on the timeline — use for dual-system sound (camera + external audio) or multicam.
manage_multicam
Create or ungroup a multicam group.
change_cam
Switch a multicam group's camera angle over timeline frame ranges, full-frame or in a multi-angle layout.
get_multicam
Read a multicam group: members (angleLabel, kind, offsetSeconds, confidence, which is master), the current program cut as run-length [angle, startFrame, endFrame) rows in timeline frames, and the track indexes the group occupies.
undo
Reverts the latest action from the editor's shared undo history, whether the user or agent made it.
get_transcript
Returns the spoken transcript of the CURRENT timeline in project frames — the post-edit caption track in one call.
remove_words
Cut speech by the word, Descript-style — the primary tool for text-based editing (filler words, flubbed sentences, dropped retakes, tightening a ramble).
remove_silence
Remove dead air — quiet, speech-free sections — from the timeline's audio, ripple-closing the gaps.
detect_beats
Detect musical beats and downbeats in a media asset's audio, on-device.
add_texts
Adds text clips as timeline layers.
update_text
Updates text clips or a captionGroupId.
add_captions
Transcribes spoken audio and creates caption text clips on their own track.
apply_color
Author/refine a color grade on video/image clips with named controls — the colorist path, distinct from apply_effect (looks/FX).
apply_effect
Apply looks and effects (FX) to clips.
inspect_color
Measure color scopes of a timeline clip's current graded look (clipId) OR a raw media asset (mediaRef) — black/white points, % clipping, mean & per-channel levels, shadow/mid/highlight color tilt, saturation, warm-cool / green-magenta balance, and a saturation-weighted hueHistogram (12 bins of...
denoise_audio
Remove background noise from audio clips using an on-device speech-enhancement model (DeepFilterNet3).
list_models
Lists AI models with their capabilities (durations, aspect ratios, resolutions, draft mode, first/last frame support, reference support, voices/category for audio, and configurable settings for upscalers).
generate_video
Starts an async AI video generation.
generate_image
Starts an async AI image generation.
generate_audio
Starts an async AI audio generation or transformation.
upscale_media
Enhances an existing video or image with an AI upscaler.
send_feedback
Report an agent limitation or bug to the Palmier team so they can improve the product.
read_skill
Load the full instructions for one of the skills listed under # Skills in your system prompt.
manage_skills
Create, update, or permanently remove reusable skills for the in-app assistant.
manage_project
List, open, create, or close Palmier projects for this MCP session.

Similar servers