Wraps the ffmpeg/ffprobe command-line tools to give agents working media capabilities: probing duration, dimensions, codecs and streams; trimming, cropping and resizing clips; extracting audio and single frames; generating thumbnail contact sheets; loudness normalization to a target LUFS; concatenating mixed-codec files; burning in SRT/ASS/VTT captions; and converting containers or codecs (with stream-copy support) — each call returning verified JSON output metadata.
Click on "Deploy Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type @ followed by the MCP server name and your instructions, e.g., "@mediakit How long is clip.mp4 and what's its resolution?"
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
Give your coding agent eyes and ears for media. MediaKit is a small MCP server + CLI that wraps ffmpeg — trim, crop, resize, captions, loudness, thumbnails, frame grabs, audio extraction, concat, and conversion — so an agent can actually work with video and audio instead of guessing.
Coding agents are blind and deaf to media. They can read a 10,000-line codebase but can't tell you how long a clip is, pull a frame from it, or make it vertical. MediaKit fixes that with one install.
npx mediakit-for-agents probe clip.mp4Or install it globally:
npm install -g mediakit-for-agents
mediakit trim clip.mp4 --start 2 --duration 8 -o clip-short.mp4Add to your MCP config (.mcp.json or your client's settings):
{
"mcpServers": {
"mediakit": {
"command": "npx",
"args": ["-y", "mediakit-for-agents", "mcp"]
}
}
}Then ask your agent things like:
"How long is this video, and what's its resolution?"
"Make this clip vertical 1080×1920 and burn the captions in."
"Normalize the voiceover to -14 LUFS and pull out the audio as MP3."
Requirements: ffmpeg and ffprobe on your PATH
(ffmpeg.org/download). Node 18+.
Related MCP server: ffmpeg-mcp
Tool | What it does | CLI |
| Duration, dimensions, codecs, streams (ffprobe) |
|
| Cut a clip by start + duration/end |
|
| Crop to a rectangle (w, h, x, y) |
|
| Resize, keeping aspect ratio by default |
|
| Pull the audio track out (mp3/aac/wav/ogg/opus/flac) |
|
| Grab one frame at a timestamp |
|
| Contact sheet of N evenly spaced frames |
|
| Normalize to a loudness target (default −14 LUFS) |
|
| Join 2+ files into one (re-encodes, mixed codecs OK) |
|
| Burn .srt/.ass/.vtt subtitles into the video |
|
| Change container/codec (omit a codec to stream-copy) |
|
Every tool returns a concise JSON result: output path, size, and the verified duration/dimensions of what it produced.
# 1. Inspect the source
mediakit probe input.mp4
# 2. Center-crop to 9:16, then resize to 1080×1920
mediakit crop input.mp4 --width 720 --height 1280 --x 280 --y 0 -o cropped.mp4
mediakit resize cropped.mp4 --width 1080 -o vertical.mp4
# 3. Burn captions on
mediakit burn-captions vertical.mp4 --captions captions.srt -o short.mp4
# 4. Check the result
mediakit probe short.mp4 # 1080x1920, same duration, captions visibleSee skills/shorts-video/SKILL.md for the
full script → captioned vertical-video workflow.
No shell is ever involved: every ffmpeg call uses execFile with an
argument array, so filenames and user input can't inject commands.
Outputs are never allowed to overwrite their inputs.
Inputs are validated (must exist and be readable) before ffmpeg runs.
skills/mediakit/SKILL.md — agent-facing
guide: when to use each tool, re-encode vs stream copy, loudness targets.
skills/shorts-video/SKILL.md — the
short-form vertical-video workflow (script → 1080×1920 captioned clip).
MIT — see LICENSE.
Burn subtitles into the video (hardcoded, always visible). Accepts .srt, .ass or .vtt. Style: white bold text on a semi-transparent black box at the bottom.
Inputs: input (required), output (required), captions (required), font_size.
Join 2+ media files into one. Re-encodes, so inputs may differ in codec or size.
Inputs: inputs (required), output (required), audio_codec, video_codec.
Change container or codec. Omit video_codec/audio_codec to stream-copy that stream. crf 23 is a good default for H.264.
Inputs: input (required), output (required), crf, preset, +2 more.
Crop a video to a rectangle. width/height/x/y in pixels; must fit inside the source frame.
Inputs: input (required), width (required), height (required), output (required), +2 more.
Extract the audio track to mp3, aac, wav, ogg, opus or flac (from the output extension).
Inputs: input (required), output (required), format.
Grab a single frame at a timestamp as PNG/JPG/WebP.
Inputs: input (required), output (required), timestamp.
Normalize loudness to a target (default -14 LUFS, the YouTube/Spotify standard) with a single-pass loudnorm filter. Good for speech and creator audio.
Inputs: input (required), output (required), format, target_lufs.
Inspect a media file with ffprobe: duration, dimensions, codecs, streams.
Inputs: input (required).
Resize a video. Give width and/or height; with keep_aspect (default) the other dimension follows the aspect ratio and is rounded to an even number.
Inputs: input (required), output (required), width, height, +1 more.
Build a contact sheet: N evenly spaced frames tiled into a grid image.
Inputs: input (required), output (required), width, frames, +1 more.
Cut a clip out of a media file. Give start (seconds) plus duration OR end. Re-encodes by default for frame-accurate cuts; set stream_copy=true for fast lossless cuts (keyframe boundaries only).
Inputs: input (required), output (required), end, start, +2 more.
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
Loading tool schema changelog…
Scored across 11 tools
Each tool targets a clearly distinct media operation: probing, converting, concatenating, trimming, cropping, resizing, audio extraction, frame extraction, thumbnails, loudness normalization, and caption burning. Overlaps such as trim vs. extract_frame and thumbnail vs. extract_frame are resolved by their descriptions and output types.
Most tools use a consistent snake_case verb_noun or verb pattern: concat, burn_captions, convert, probe, trim, crop, resize, extract_audio, extract_frame. The nouns thumbnail and loudness are minor deviations, though they remain readable and predictable.
Eleven tools is well-scoped for a media processing toolkit, covering common ffmpeg workflows without excessive fragmentation. Each tool has a clear job and none feels redundant.
The surface covers most core media lifecycle operations: inspect, convert, trim, concat, crop, resize, extract audio/frames, thumbnail, loudness, and captions. Minor gaps exist, such as replacing or muxing separate audio/video tracks and extracting embedded subtitles, but agents can work around many of these with existing tools.
Video, audio, and image processing for AI agents: convert, transcribe, upscale - 150+ operations.
Transform video, audio and images, and generate media from prompts. FFmpeg, captions, models.
FFmpeg as a service for AI agents: typed video editing tools, async jobs, downloadable outputs.
PDF, image, video, OCR, screenshot, SQL, QR and text tools for agents. No API key, no signup.
