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., "@lazy-media-mcp prepare /Users/me/demo.mp4 for AI vision"
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.
Local Model Context Protocol (MCP) server that compresses images/videos and prepares media for AI vision agents.
Designed for coding agents (Claude Code, Codex, Grok, and other MCP clients): returns file paths only (no inline base64), so large screenshots and demos stay within tool limits.
Keywords: MCP server, image compression, video compression, ffmpeg, sharp, AI vision prep, frame extraction, Claude, Codex.
Large screenshots and long demos burn context and often fail tool limits. This server:
Shrinks images to a sensible size/quality
Turns videos into frame packs agents can actually open
Uses JPEG by default for widest agent compatibility
Format | Role | Default here? |
JPEG | Best universal image input for local agents | Yes |
PNG | Sharper for OCR / UI text / alpha |
|
WebP | Smaller files when the host supports it | Opt-in only |
MP4 | Storage/sharing re-encode | Video compress default |
WebM | Optional container | Opt-in via |
Format does not improve model understanding by itself. Resolution, blur, and compression artifacts matter more. Over-aggressive WebP/JPEG hurts OCR.
Local agents usually do not natively watch WebM/MP4. Prefer prepare_for_ai / video_extract_frames → JPEG paths.
Related MCP server: mcp-media-engine
Node.js ≥ 20
ffmpeg + ffprobe on PATH (video tools)
# macOS
brew install ffmpegnpx -y lazy-media-mcp
# or
npm install -g lazy-media-mcpgit clone https://github.com/leaf76/lazy-media-mcp.git
cd lazy-media-mcp
npm install
npm run build
npm test
node dist/cli.js # stdio MCP{
"mcpServers": {
"lazy-media": {
"command": "node",
"args": ["/absolute/path/to/lazy-media-mcp/dist/cli.js"],
"env": {
"MEDIA_ALLOWED_ROOTS": "/Users/you,/Users/you/WorkSpace",
"MEDIA_WORKDIR": "/Users/you/.cache/lazy-media-mcp/jobs"
}
}
}
}Tool | Purpose |
| Metadata only |
| Resize/compress image → workdir path |
| Re-encode video (default MP4) |
| Extract frames for vision |
| One-shot profile pipeline (recommended) |
| Delete a job directory by |
prepare_for_ai profilesProfile | Behavior |
| Image → JPEG ≤1536 edge; video → up to 10 JPEG frames |
| Prefer PNG / higher quality |
| Smaller edges, fewer frames |
| Higher quality + optional compressed MP4 |
Variable | Default |
|
|
|
|
| 500MB |
| 200MB |
| 24 |
| 120000 |
|
|
|
|
Path allowlist (realpath checks)
Input/output size caps
Process timeout
ffmpeg/ffprobe invoked with argv arrays only (no shell interpolation)
Outputs go to workdir; originals are not overwritten
Cleanup only deletes direct children of workdir by job_id
1. prepare_for_ai({ path: "/path/to/demo.mp4", profile: "ai_vision" })
2. Read returned outputs[].path frame files in the next vision step
3. media_cleanup({ job_id }) when done (optional)lazy-desktop-mcp — desktop computer-use MCP
lazy_mobile_mcp — Android/iOS mobile automation MCP
session-collab-mcp — multi-session collaboration MCP
MIT © leaf76
Compress/resize an image to a job workdir. Default format is JPEG for local agent compatibility. Returns output file paths only.
Inputs: path (required), format, job_id, quality, +2 more.
Delete a previous job directory under the media workdir by job_id.
Inputs: job_id (required).
Inspect image or video metadata (dimensions, codecs, duration, size). Does not modify files.
Inputs: path (required).
One-shot AI prep: images are resized/compressed; videos become a frame pack (default). Paths only — no inline base64.
Inputs: path (required), job_id, profile.
Re-encode a video (default MP4/H.264). WebM is optional. Prefer video_extract_frames or prepare_for_ai for agent vision.
Inputs: path (required), crf, job_id, container, +2 more.
Extract frames from a video into JPEG/PNG/WebP files for AI vision agents. Returns file paths only.
Inputs: path (required), mode, format, job_id, +4 more.
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
Loading tool schema changelog…
Scored across 6 tools
Each tool has a clearly distinct purpose: AI preparation, metadata inspection, compression for images and videos separately, frame extraction, and cleanup. No overlapping functionality.
Most tools follow a noun_verb pattern (media_inspect, image_compress, video_compress, video_extract_frames, media_cleanup), but prepare_for_ai breaks the pattern with a verb_preposition_noun structure. Still, all use snake_case and are readable.
With 6 tools, the server is well-scoped for media processing and AI preparation. Each tool addresses a specific need without redundancy or excessive granularity.
Covers core operations (compression, inspection, frame extraction, cleanup, and one-shot AI prep). Missing a list_jobs tool to retrieve previous job IDs, but the main workflow is supported.
OCR, transcription, file extraction, and image generation for AI agents via MCP.
MCP server for Clipkit — gives AI agents a video toolbox via the Clipkit schema.
MCP server for Wan AI video generation
MCP server for QPost — lets AI agents publish video and image posts to YouTube, TikTok, Instagram.
