Agent Skills

ffmpeg-video-editor

A complete FFmpeg-powered video editing agent. Use this skill whenever the user wants to do anything with video or audio files — trimming, cutting, transcoding, resizing, concatenating, adding subtitles, overlaying images or text, applying color grades or LUTs, normalizing audio levels, creating transitions, green-screen compositing, frame extraction, slideshows, or exporting for social media (TikTok, Reels, YouTube). Also use it for codec questions (H.264, H.265, VP9, AV1, ProRes, FFV1), hardwa

Install

npx skills add https://github.com/bryanwhl/ffmpeg-video-editor --skill ffmpeg-video-editor
SKILL.md

FFmpeg Video Editor Agent Skill

You are an expert FFmpeg operator. Use this skill to answer any video/audio editing request by constructing correct ffmpeg command lines. Always probe input files first, then select the appropriate operation category below.

Probe Input Files

Run scripts/probe.sh <file> (make executable first with chmod +x scripts/*.sh if needed), or inspect manually:

ffprobe -v error -show_streams -show_format -of json input.mp4

Trim / Cut

# Fast stream-copy trim (no re-encode)
ffmpeg -ss 00:01:00 -to 00:02:30 -i input.mp4 -c copy out.mp4

# Precise trim (re-encode around keyframes)
ffmpeg -ss 00:01:00 -i input.mp4 -t 90 -c:v libx264 -crf 18 -c:a aac out.mp4

Transcode

# H.264 web-safe (pix_fmt ensures broad player compatibility)
ffmpeg -i input.mp4 -c:v libx264 -crf 23 -preset slow \
  -pix_fmt yuv420p -c:a aac -b:a 128k -movflags +faststart out.mp4

# H.265 (smaller files)
ffmpeg -i input.mp4 -c:v libx265 -crf 28 -preset slow \
  -pix_fmt yuv420p -c:a aac -b:a 128k out.mp4

# VP9 for web
ffmpeg -i input.mp4 -c:v libvpx-vp9 -crf 33 -b:v 0 \
  -c:a libopus -b:a 128k out.webm

# AV1 (best compression)
ffmpeg -i input.mp4 -c:v libaom-av1 -crf 35 -cpu-used 4 \
  -c:a libopus -b:a 128k out.mkv

See references/CODECS.md for full codec comparison.

Concatenate Files

# Using helper script
scripts/concat.sh out.mp4 clip1.mp4 clip2.mp4 clip3.mp4

# Manual concat demuxer
printf "file 'a.mp4'\nfile 'b.mp4'\n" > list.txt
ffmpeg -f concat -safe 0 -i list.txt -c copy out.mp4

Scale / Resize

# Scale to 1920×1080, keep aspect ratio, pad to fill
ffmpeg -i in.mp4 -vf "scale=1920:1080:force_original_aspect_ratio=decrease,\
pad=1920:1080:(ow-iw)/2:(oh-ih)/2:black" out.mp4

# Scale by width only (auto height)
ffmpeg -i in.mp4 -vf "scale=1280:-2" out.mp4

Crop

# Crop 640×360 starting at (100,50)
ffmpeg -i in.mp4 -vf "crop=640:360:100:50" out.mp4

# Center crop to 1:1 square
ffmpeg -i in.mp4 -vf "crop=min(iw\,ih):min(iw\,ih)" out.mp4

Overlay / Watermark

# Overlay image (top-right, 10px margin)
ffmpeg -i video.mp4 -i logo.png \
  -filter_complex "[0:v][1:v] overlay=W-w-10:10" out.mp4

# PiP: small video on top of main
ffmpeg -i main.mp4 -i pip.mp4 \
  -filter_complex "[1:v]scale=320:-2[small];[0:v][small]overlay=W-w-10:H-h-10" out.mp4

Draw Text / Titles

ffmpeg -i in.mp4 -vf \
  "drawtext=text='My Title':fontfile=/usr/share/fonts/truetype/dejavu/DejaVuSans.ttf\
:fontsize=48:fontcolor=white:x=(w-text_w)/2:y=(h-text_h)/2\
:shadowx=2:shadowy=2" out.mp4

Burn Subtitles

# SRT subtitles (re-encode required)
ffmpeg -i video.mp4 -vf "subtitles=subs.srt" out.mp4

# ASS/SSA subtitles (preserves styling)
ffmpeg -i video.mp4 -vf "ass=subs.ass" out.mp4

Speed Change

# 2× speed (video + audio)
ffmpeg -i in.mp4 -vf "setpts=0.5*PTS" -af "atempo=2.0" out.mp4

# 0.5× speed (half speed)
ffmpeg -i in.mp4 -vf "setpts=2.0*PTS" -af "atempo=0.5" out.mp4

# >4×: chain atempo (max 2.0 per filter)
ffmpeg -i in.mp4 -vf "setpts=0.25*PTS" -af "atempo=2.0,atempo=2.0" out.mp4

# <0.5×: chain atempo (min 0.5 per filter)
ffmpeg -i in.mp4 -vf "setpts=4.0*PTS" -af "atempo=0.5,atempo=0.5" out.mp4

Reverse

ffmpeg -i in.mp4 -vf "reverse" -af "areverse" out.mp4

Transitions (xfade)

# Dissolve between two clips (1s overlap at t=5)
ffmpeg -i a.mp4 -i b.mp4 \
  -filter_complex "[0:v][1:v]xfade=transition=dissolve:duration=1:offset=5[v];\
[0:a][1:a]acrossfade=d=1[a]" \
  -map "[v]" -map "[a]" out.mp4

See references/TRANSITIONS.md for all 40+ xfade transition types.

Audio: Volume & Normalization

# Adjust volume
ffmpeg -i in.mp4 -af "volume=1.5" out.mp4

# EBU R128 normalization (two-pass)
scripts/normalize-audio.sh in.mp4 out.mp4

# One-pass loudnorm
ffmpeg -i in.mp4 -af "loudnorm=I=-16:TP=-1.5:LRA=11" out.mp4

Audio: Fade In/Out

# Fade in 2s at start, fade out 3s before end (total 60s)
ffmpeg -i in.mp4 -af "afade=t=in:d=2,afade=t=out:st=57:d=3" out.mp4

Mix / Combine Audio

# Mix two audio streams
ffmpeg -i video.mp4 -i music.mp3 \
  -filter_complex "[0:a][1:a]amix=inputs=2:duration=first:weights=1 0.3[a]" \
  -map 0:v -map "[a]" -c:v copy out.mp4

Color Grading

# Brightness/Contrast/Saturation
ffmpeg -i in.mp4 -vf "eq=brightness=0.05:contrast=1.2:saturation=1.3" out.mp4

# Hue shift
ffmpeg -i in.mp4 -vf "hue=h=30:s=1.1" out.mp4

# Apply 3D LUT (.cube)
ffmpeg -i in.mp4 -vf "lut3d=file=lut.cube" out.mp4

# Apply Hald CLUT PNG
ffmpeg -i in.mp4 -i lut.png \
  -filter_complex "[0:v][1:v]haldclut" out.mp4

Chroma Key / Green Screen

# chromakey (color + similarity + blend)
ffmpeg -i foreground.mp4 -i background.mp4 \
  -filter_complex "[0:v]chromakey=0x00FF00:0.3:0.2[fg];\
[1:v][fg]overlay" out.mp4

# colorkey (simpler, exact color)
ffmpeg -i in.mp4 -vf "colorkey=green:0.3:0.2" out.mp4

Frame Rate Conversion

# Convert to 30fps
ffmpeg -i in.mp4 -vf "fps=30" out.mp4

# Use motion interpolation (minterpolate)
ffmpeg -i in.mp4 -vf "minterpolate=fps=60:mi_mode=mci" out.mp4

Extract Frames / Create Slideshow

# Extract one frame per second as PNG
ffmpeg -i in.mp4 -vf "fps=1" frames/%04d.png

# Slideshow from images (3s per image, 30fps output)
ffmpeg -framerate 1/3 -i frames/%04d.png -c:v libx264 -r 30 out.mp4

Stacking / Grid Layout

# Side by side (hstack)
ffmpeg -i a.mp4 -i b.mp4 -filter_complex "[0:v][1:v]hstack=inputs=2" out.mp4

# Top-bottom (vstack)
ffmpeg -i a.mp4 -i b.mp4 -filter_complex "[0:v][1:v]vstack=inputs=2" out.mp4

# 2×2 grid (xstack)
ffmpeg -i a.mp4 -i b.mp4 -i c.mp4 -i d.mp4 \
  -filter_complex "[0:v][1:v][2:v][3:v]xstack=inputs=4:layout=0_0|w0_0|0_h0|w0_h0" out.mp4

Hardware-Accelerated Encoding

# NVIDIA NVENC H.264
ffmpeg -hwaccel cuda -i in.mp4 -c:v h264_nvenc -preset p4 -cq 23 out.mp4

# Apple VideoToolbox H.265
ffmpeg -i in.mp4 -c:v hevc_videotoolbox -q:v 60 out.mp4

# Intel QuickSync H.264
ffmpeg -hwaccel qsv -i in.mp4 -c:v h264_qsv -global_quality 23 out.mp4

See references/HARDWARE-ACCEL.md for full GPU encoding guide.

Encoding Presets (assets/preset-profiles.json)

When the task maps to a named use case below, read assets/preset-profiles.json and use the example field as the ready-made ffmpeg command.

Profile Codec CRF Use Case
web-optimized H.264 23 Browser streaming
high-quality H.264 18 Archive / master
social-media-vertical H.264 23 9:16 TikTok/Reels
social-media-square H.264 23 1:1 Instagram
fast-preview H.264 28 Proxy / review
lossless-intermediate FFV1 — Post-production store

Multi-Stream Filter Graph Pattern

ffmpeg \
  -i input_video.mp4 \
  -i overlay.png \
  -i audio_bed.mp3 \
  -filter_complex "
    [0:v]scale=1920:1080,setsar=1[base];
    [1:v]scale=200:-2,format=rgba,colorchannelmixer=aa=0.8[logo];
    [base][logo]overlay=W-w-20:20[composited];
    [0:a][2:a]amix=inputs=2:weights=1 0.2[audio_mix]
  " \
  -map "[composited]" -map "[audio_mix]" \
  -c:v libx264 -crf 18 -preset slow \
  -c:a aac -b:a 192k \
  -movflags +faststart \
  output.mp4

Common Options Reference

Flag Meaning
-ss Seek/start time (before -i = fast seek)
-to End time (absolute)
-t Duration
-c copy Stream copy (no re-encode)
-crf Constant Rate Factor (lower = better)
-preset Encoder speed/quality tradeoff
-movflags +faststart Move moov atom to front (web streaming)
-vf Video filter chain
-af Audio filter chain
-filter_complex Multi-input/output filter graph
-map Select output streams explicitly
-an Drop audio
-vn Drop video
-y Overwrite output without prompt

Debugging Tips

# Dry-run: print filter graph without encoding
ffmpeg -i in.mp4 -vf "scale=1280:-2" -f null -

# Benchmark encode speed
ffmpeg -benchmark -i in.mp4 -c:v libx264 -f null -

# Show available encoders
ffmpeg -encoders | grep -E "^.V|^.A"

# Show filter options
ffmpeg -help filter=loudnorm

Related skills

video-editgenmedia-labs715KEdit existing video on RunComfy — this skill is a smart router that matches the user's intent to the right edit model in the RunComfy catalog. Picks Wan 2.7 Edit-Video (general restyle / background swap / packaging swap, identity + motion preservation), Kling 2.6 Pro Motion Control (transfer precise motion from a reference video to a target character), or Lucy Edit Restyle (lightweight identity-stable restyle / outfit swap). Bundles each model's documented prompting patterns so the skill gets shai-video-generationgenmedia-labs714KGenerate AI videos on RunComfy via the `runcomfy` CLI — a smart router across the full video-model catalog: HappyHorse 1.0 (Arena #1, native in-pass audio), Wan-AI Wan 2-7 (open weights, audio-driven lip-sync), ByteDance Seedance v2 / 1-5 / 1-0 (multi-modal cinematic), Kling 3.0 / 2-6, Google Veo 3-1, MiniMax Hailuo 2-3, ByteDance Dreamina 3-0. Covers text-to-video (t2v), image-to-video (i2v), and Veo's video-extend endpoint. The skill picks the right model for the user's intent (Arena-#1 qualitai-musicgenmedia-labs714KGenerate AI music on RunComfy via the `runcomfy` CLI — a smart router across the music-model catalog. Routes to ElevenLabs AI Music Generation (premium 44.1 kHz stereo vocal tracks, 5 s–5 min, $0.0083/s) and ACE Step / ACE Step 1.5 (StepFun-AI open-weights, tag-driven composition, multilingual lyrics, $0.0002–0.0003/s, ~27× cheaper), plus ACE Step audio-inpaint (regenerate a time range inside an existing track) and ACE Step audio-outpaint (extend a track before or after). Picks the right model fimage-to-videogenmedia-labs713KAnimate any still image on RunComfy — this skill is a smart router that matches the user's intent to the right i2v model in the RunComfy catalog. Picks HappyHorse 1.0 I2V (Arena #1, native audio, identity preservation) for general animations, Wan 2.7 with `audio_url` for custom-voiceover lip-sync, or Seedance 2.0 Pro for multi-modal animation from image + reference video + reference audio. Bundles each model's documented prompting patterns so the caller gets sharper output without burning iterat

Search skills and MCP servers

Fuzzy search across 23,137 skills and servers