AgentSkillsCN

video-production

全流程管理 Hedge Edge 的视频制作流水线:从脚本构思、MT5/cTrader/Electron 应用的录屏流程,到基于 FFmpeg 的剪辑与后期处理,再到通过 Canva 设计缩略图、优化音频、生成字幕,以及批量生产流程。适用于任何视频内容的制作、编辑或后期处理时使用。

SKILL.md
--- frontmatter
name: video-production
description: |
  Manages the full video production pipeline for Hedge Edge: scripting structure,
  screen recording workflows for MT5/cTrader/Electron app, FFmpeg-based editing
  and processing, thumbnail design via Canva, audio optimization, caption
  generation, and batch production workflows. Use when producing, editing, or
  post-processing any video content.

Video Production

Objective

Produce professional-quality video content that demonstrates Hedge Edge's product, teaches hedging concepts, and builds trust with prop firm traders all at a sustainable production pace. Maintain a batch-production workflow that generates 4-8 polished videos per cycle with minimal per-video effort through templates, automation, and reusable assets.

When to Use This Skill

  • When setting up a screen recording session for MT5, cTrader, or the Hedge Edge Electron app
  • When editing raw footage into a polished video using FFmpeg
  • When creating thumbnails for YouTube or social media via Canva
  • When extracting clips from long-form videos for Shorts or Reels
  • When adding captions, text overlays, or visual effects to video content
  • When optimizing audio (noise reduction, level normalization, music mixing)
  • When batch-producing multiple videos in a single production cycle
  • When converting video formats for different platforms (horizontal vertical, bitrate adjustments)
  • When creating animated diagrams, equity curve visualizations, or other explanatory visuals

Input Specification

InputTypeRequiredDescription
actionstringYesrecord-setup, edit, thumbnail, extract-clip, add-captions, optimize-audio, batch-produce, convert-format, create-visual
source_video_pathstringNoPath to source video file (required for editing actions)
output_pathstringNoDesired output file path. Default: auto-generated in project output directory
video_typestringNotutorial, walkthrough, explainer, shorts, reel, testimonial, announcement
thumbnail_textstringNoText to overlay on thumbnail (for thumbnail action)
thumbnail_stylestringNodramatic, clean, data-focused, before-after. Default: dramatic
clip_startstringNoStart timestamp for clip extraction, format: HH:MM:SS or seconds
clip_endstringNoEnd timestamp for clip extraction
caption_filestringNoPath to SRT/VTT caption file (for add-captions)
target_platformstringNoyoutube, youtube-shorts, instagram-reel, instagram-post, linkedin
batch_itemslist[object]NoList of video items for batch production, each with script reference and assets

Step-by-Step Process

Phase 1: Recording Setup

Screen recording configuration for trading software demos:

  1. MT5/MetaTrader 5 recording setup:

    • Resolution: 1920x1080 (full HD), 60fps for smooth chart animation
    • Clean the MT5 workspace: remove unnecessary toolbars, set chart to clean template
    • Load Hedge Edge EA onto a demo account with visible trade history
    • Prepare demo scenario: pre-planned trades to execute during recording
    • Display key elements: trade journal, equity curve, account balance, open positions
    • Use a consistent MT5 color scheme that matches Hedge Edge branding (dark theme preferred)
  2. Hedge Edge Electron app recording:

    • Resolution: 1920x1080, 60fps
    • Show full app window with all panels visible
    • Prepare test accounts: at least one prop account + one hedge account connected
    • Stage a live demonstration: open a position on prop account, show instant hedge execution
    • Highlight key UI elements: account overview, trade log, hedge status indicators
  3. Webcam/presenter recording (optional overlay):

    • 720p webcam feed, positioned bottom-right corner
    • Clean background or green screen
    • Good lighting (ring light or key/fill setup)
    • Consistent framing (chest up, centered)
  4. Audio setup:

    • External microphone preferred (USB condenser or lapel mic)
    • Record at 48kHz sample rate, 16-bit
    • Record in quiet environment; use noise gate if needed
    • Keep audio and video in sync (clap sync or software sync)

Recording software recommendations:

  • OBS Studio (free, screen + webcam composite)
  • Windows Game Bar (quick screen capture)
  • Separate audio recording via Audacity for higher quality

Phase 2: Video Editing with FFmpeg

Standard editing pipeline:

  1. Import and organize: raw/ screen-recording.mp4 webcam.mp4 audio.wav assets/ intro-animation.mp4 outro-animation.mp4 lower-third.png logo-watermark.png

  2. Basic assembly (FFmpeg commands):

    Trim footage: ffmpeg -i input.mp4 -ss 00:00:30 -to 00:15:00 -c copy trimmed.mp4

    Combine screen recording + webcam overlay: ffmpeg -i screen.mp4 -i webcam.mp4 -filter_complex "[1:v]scale=320:240[pip];[0:v][pip]overlay=W-w-20:H-h-20" -c:a copy output.mp4

    Add intro/outro: ffmpeg -f concat -i filelist.txt -c copy assembled.mp4 (filelist.txt contains: file 'intro.mp4' / file 'main.mp4' / file 'outro.mp4')

    Add logo watermark: ffmpeg -i input.mp4 -i logo.png -filter_complex "overlay=W-w-20:20:enable='between(t,5,999)'" -c:a copy watermarked.mp4

    Normalize audio levels: ffmpeg -i input.mp4 -af loudnorm=I=-16:TP=-1.5:LRA=11 -c:v copy normalized.mp4

  3. Advanced editing:

    Speed ramp (speed up boring sections): ffmpeg -i input.mp4 -filter:v "setpts=0.5*PTS" -filter:a "atempo=2.0" fast.mp4

    Picture-in-picture with rounded corners: ffmpeg -i bg.mp4 -i pip.mp4 -filter_complex "[1:v]scale=400:225,format=yuva420p,geq=lum='p(X,Y)':a='if(gt(abs(X-W/2),W/2-20)*gt(abs(Y-H/2),H/2-20),0,255)'[pip];[0:v][pip]overlay=50:50" output.mp4

    Add background music (low volume under voice): ffmpeg -i voice.mp4 -i music.mp3 -filter_complex "[1:a]volume=0.15[bg];[0:a][bg]amix=inputs=2:duration=first" -c:v copy mixed.mp4

Phase 3: Clip Extraction for Shorts/Reels

  1. Identify clip-worthy moments from the full video:

    • The single most compelling 30-60 second segment
    • A surprising stat or revelation
    • A visual "aha moment" (e.g., watching the hedge execute in real-time)
    • A controversial or contrarian take that stands alone
  2. Extract and reformat:

    Extract clip: ffmpeg -i full-video.mp4 -ss 00:05:30 -to 00:06:15 -c copy clip.mp4

    Convert to vertical (9:16) for Shorts/Reels: ffmpeg -i clip.mp4 -vf "scale=1080:1920:force_original_aspect_ratio=decrease,pad=1080:1920:(ow-iw)/2:(oh-ih)/2:black" vertical.mp4

    Add top/bottom text bars for context: ffmpeg -i vertical.mp4 -vf "drawtext=text='How to Hedge a $100K FTMO Challenge':fontsize=40:fontcolor=white:x=(w-text_w)/2:y=100:fontfile=arial.ttf" captioned.mp4

  3. Shorts/Reels specifications:

    • Duration: 15-60 seconds (sweet spot: 30-45 seconds)
    • Aspect ratio: 9:16 (1080x1920)
    • Hook within first 2 seconds
    • Auto-captions always enabled (80%+ viewers watch muted)
    • End with a clear next step: "Follow for more" or "Full video linked above"

Phase 4: Thumbnail Design via Canva

  1. Thumbnail principles for trading content:

    • High contrast: Dark background with bright text and elements
    • Readable at mobile size: Test at 168x94px (YouTube mobile thumbnail)
    • 3-element rule: Face/reaction (optional), text headline, visual proof (chart/screenshot)
    • Emotional trigger: Curiosity, fear of missing out, or problem recognition
    • Brand consistency: Hedge Edge logo in corner, consistent color palette
  2. Thumbnail templates (via Canva API):

    Template A "Dramatic Result":

    • Dark gradient background (Hedge Edge brand colors)
    • Large bold text: "I HEDGED MY $100K FTMO CHALLENGE"
    • Equity curve screenshot showing before/after
    • Hedge Edge logo bottom-right

    Template B "Step-by-Step":

    • Clean dark background
    • Numbered steps visible: "3 Steps to..."
    • MT5 screenshot with highlighted UI elements
    • Arrow or circle annotations pointing to key areas

    Template C "Versus/Comparison":

    • Split layout: left (red/loss) vs right (green/profit)
    • "WITHOUT Hedge Edge" vs "WITH Hedge Edge"
    • Contrasting equity curves or P&L numbers
    • Bottom text: prop firm names (FTMO, The5%ers)

    Template D "Data/Stats":

    • Central large number or stat ("85% FAIL")
    • Supporting text below
    • Chart or data visualization in background
    • Clean, professional, data-journalism aesthetic
  3. Canva API workflow:

    • Authenticate with CANVA_API_KEY
    • Select base template by thumbnail style
    • Inject dynamic text (video title, stats)
    • Inject screenshot/image assets
    • Export at 1280x720px, JPG, <2MB
    • A/B test 2 thumbnails per video when possible

Phase 5: Audio Optimization

  1. Noise reduction: ffmpeg -i input.mp4 -af "afftdn=nf=-25" denoised.mp4

  2. Loudness normalization (YouTube standard: -14 LUFS): ffmpeg -i input.mp4 -af loudnorm=I=-14:TP=-1.5:LRA=11 normalized.mp4

  3. De-essing (reduce sibilance): ffmpeg -i input.mp4 -af "equalizer=f=6000:t=q:w=1:g=-6" deessed.mp4

  4. Background music mixing:

    • Music volume: -18dB below voice
    • Fade in over 3 seconds at intro
    • Fade out over 3 seconds at outro
    • Duck under voice sections, slightly louder during transitions

Phase 6: Caption Generation

  1. Generate captions:

    • Use OpenAI Whisper API or local Whisper model for speech-to-text
    • Output format: SRT (YouTube) or VTT (web)
    • Review and correct trading-specific terms: "FTMO", "MetaTrader", "Hedge Edge", "drawdown", "pip"
  2. Burn captions into video (for Shorts/Reels where platform captions aren't available): ffmpeg -i input.mp4 -vf "subtitles=captions.srt:force_style='FontName=Arial,FontSize=24,PrimaryColour=&H00FFFFFF,OutlineColour=&H00000000,Outline=2,Alignment=2'" captioned.mp4

  3. Caption styling for Shorts/Reels:

    • Centered, bottom third of frame
    • White text with black outline (maximum readability)
    • 2-3 words per line maximum
    • Highlight key terms in brand accent color

Phase 7: Batch Production Workflow

  1. Batch cycle (recommended biweekly):

    • Day 1: Script 4-6 videos (use content-creation skill)
    • Day 2: Record all screen captures and voiceovers in one session
    • Day 3-4: Edit all videos (assembly, overlays, audio)
    • Day 5: Create all thumbnails, extract all shorts clips
    • Day 6: Review, caption, and finalize all videos
    • Day 7: Upload all to YouTube (scheduled), schedule all derivatives
  2. Batch FFmpeg processing (process all videos with consistent settings): for f in raw/*.mp4; do ffmpeg -i "$ " -af loudnorm=I=-14:TP=-1.5:LRA=11 \ -vf "movie=logo.png[wm];[in][wm]overlay=W-w-20:20" \ "processed/$(basename $ )" done

  3. Asset reuse:

    • Intro/outro animations: same across all videos (update quarterly)
    • Lower-third template: consistent name/title card
    • Music library: 5-10 approved tracks rotated across videos
    • Transition effects: standardized set (cut, dissolve, wipe)

Output Specification

  • Record Setup: Returns detailed recording configuration checklist with software settings, demo scenario plan, and asset requirements
  • Edit: Returns processed video file at output path with edit log (operations performed, duration, resolution)
  • Thumbnail: Returns Canva design URL and exported thumbnail file (1280x720 JPG)
  • Extract Clip: Returns vertical clip file ready for Shorts/Reels upload with duration and format confirmation
  • Add Captions: Returns captioned video file and SRT/VTT file for platform upload
  • Optimize Audio: Returns audio-optimized video with loudness report (LUFS measurement, noise floor)
  • Batch Produce: Returns production summary: videos completed, total duration, thumbnails created, clips extracted, and next steps
  • Convert Format: Returns converted video file with format specification confirmation
  • Create Visual: Returns animated/static visual asset (equity curve animation, diagram, data visualization)

API & Platform Requirements

PlatformAPI/ToolEnv VariablePurpose
FFmpegFFmpeg CLIN/A (local install)Video editing, audio processing, format conversion, clip extraction, caption burning
CanvaCanva APICANVA_API_KEYThumbnail design, visual templates, brand asset management
OpenAIOpenAI APIOPENAI_API_KEYWhisper transcription for captions, script assistance
OBS StudioOBS CLI/APIN/A (local install)Screen recording automation

Quality Checks

  • All videos exported at minimum 1080p (1920x1080) at 30fps or higher
  • Audio normalized to -14 LUFS for YouTube, -16 LUFS for social platforms
  • Thumbnails are 1280x720px, under 2MB, and readable at mobile thumbnail size (168x94px)
  • Every Shorts/Reels clip has burned-in captions and a hook within the first 2 seconds
  • Hedge Edge logo watermark present on all video content (after intro, persistent)
  • No copyrighted music used only royalty-free or licensed tracks from approved library
  • Screen recordings show clean MT5/app interfaces with no personal data, real account numbers, or sensitive information visible
  • Vertical content is properly formatted at 9:16 (1080x1920) with no letterboxing artifacts
  • Batch production cycles produce at least 4 publish-ready videos per cycle
  • All raw footage is backed up before editing begins
  • Trading terminology is correctly spelled in all captions (FTMO, MetaTrader, cTrader, Hedge Edge never misspelled)
  • Color grading is consistent across all videos in a batch (same LUT or color profile)