nira-video-mcp 1.0.1

NIRA Video Production MCP — professional video editing, grading, AI generation, and delivery (51 tools)

NIRA Video Production MCP

The most complete video production MCP for Claude. 51 tools covering the full post-production pipeline — editing, color grading, VFX, motion graphics, audio mastering, AI video generation, and platform delivery.

Features

Inspection — probe metadata, extract thumbnails, visualize waveforms
Editing — trim, concat, resize, speed, stabilize, transitions (57+ xfade types)
Color Grading — 8 cinematic LUTs (teal/orange, vintage film, noir, blockbuster...), custom .cube LUT support, grain, vignette, chromakey, glitch/VHS, halation
Audio — normalize (EBU R128 two-pass), mix, replace, background music with auto-ducking, silence removal
Motion Graphics — text overlays with animation, lower thirds, title cards, subtitle burn-in, progress bars, Manim mathematical animations
AI Generation — local GPU generation via Wan2GP (free, private) or cloud via fal.ai (Kling 3.0, MiniMax Hailuo, Luma Ray3, Veo 3.1, Pika 2.2)
Delivery — 15 platform presets (YouTube 4K/1080/Shorts, TikTok, Instagram Reels/Feed, Twitter, Facebook, LinkedIn, ProRes, H.265, AV1, HLS adaptive streaming, GIF)
Projects — multi-clip project management with sequential xfade renders and background music
Render Queue — all heavy operations run as background jobs; poll status without blocking

Requirements

• Python 3.11+
• FFmpeg in PATH (includes ffprobe)
• Claude Code
• pip install "mcp[cli]>=1.27.0" pillow fal-client

Optional:

• Wan2GP for free local AI video generation
• fal.ai API key for cloud AI generation (Kling, Luma, Veo, etc.)

Installation

git clone https://github.com/pachote/nira-video-mcp
cd nira-video-mcp
pip install "mcp[cli]>=1.27.0" pillow fal-client

Register with Claude Code:

claude mcp add nira-video-prod python server.py \
  --env VIDPROD_OUTPUT="/path/to/output" \
  --env VIDPROD_INPUT="/path/to/clips" \
  --env FAL_KEY="your-fal-key-here"

Or add to .mcp.json — see .mcp.json.example.

Environment Variables

Variable	Default	Description
VIDPROD_WORKSPACE	~/video_workspace	Working directory for projects
VIDPROD_INPUT	~/video_input	Input clips directory
VIDPROD_OUTPUT	~/video_output	Rendered output directory
FAL_KEY	—	fal.ai API key for cloud AI generation
WAN2GP_PATH	D:/Wan2GP	Path to Wan2GP installation

Usage Examples

Edit and export for YouTube:

"Trim my_clip.mp4 to the first 60 seconds, apply the teal_orange grade, normalize audio to -16 LUFS, export for YouTube 1080p"

AI video generation:

"Generate a 5-second cinematic aerial shot of a neon city at night using Kling 3.0"

Full project pipeline:

"Create a project from clips A, B, C. Add fade transitions, background music at 20% volume, add a lower third with my name, export for Instagram Reels"

Motion graphics:

"Add an animated lower third 'John Smith — CEO' starting at 5 seconds"

All 51 Tools

Inspection

vid_inspect vid_probe vid_thumbnail vid_list_workspace vid_waveform

Editing

vid_trim vid_concat vid_resize vid_speed vid_filter vid_overlay vid_loop vid_reverse vid_stabilize vid_transition

Color Grading & VFX

vid_grade vid_lut vid_grain vid_vignette vid_chromakey vid_glitch vid_halation vid_motion_blur

Audio

aud_extract aud_replace aud_normalize aud_mix aud_add_music aud_silence

Motion Graphics

vid_text_overlay vid_lower_third vid_title_card vid_subtitles vid_progress_bar vid_manim_render

AI Generation

ai_gen_wan2gp ai_gen_fal ai_gen_status ai_gen_cancel ai_gen_list_models

Delivery & Export

vid_export vid_export_hls vid_thumbnail_sprite vid_chapters vid_metadata

Projects

proj_create proj_list proj_render

Render Queue

render_status render_cancel render_list

License

MIT