agentleFS
Sign inStart free
rahulcommercialrahulcommercial/claude-code-for-content-creators/skills/reels-pipeline/SKILL.md

reels-pipeline

Skill0 starsChanged 5 months agoMITView on GitHub

Transforms a raw landscape clip into a post-ready 9:16 Reel — crop/pad, BGM mixing, caption overlay, hook frame, end card. Wraps FFmpeg in a one-shot pipeline. Trigger when user has a horizontal clip and wants a 9:16 Reel/Short, or mentions Reels pipeline, vertical video, IG Reels, YT Shorts.

The file

We split this skill into the rules it gives an agent

3 rules. About 0 are common, 3 are distinctive, and 0 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈3

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 0

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Don't centre-crop a horizontal gameplay clip blindly — kill-feed and minimap get lost.

    line 64
  2. Distinctive:

    Don't mix BGM louder than -15 LUFS — voice gets buried.

    line 65
  3. Distinctive:

    Don't burn captions on the bottom 10% — IG UI covers it.

    line 66

≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Reels Pipeline

Raw 16:9 → polished 9:16 in one pass.

When to use

  • User has a single source clip (gameplay, screen recording, talking head).
  • Target is IG Reels / YT Shorts / TikTok (1080×1920, ≤ 90s).
  • User wants captions burned in (better retention than IG's auto-captions).

Pipeline stages

input.mp4 (16:9)
   ↓ 1. trim to hook + body + payoff (≤ 60s)
   ↓ 2. crop or pad to 9:16
   ↓ 3. mix BGM at -18 LUFS under voice
   ↓ 4. burn captions (SRT → ass with style)
   ↓ 5. prepend 1.5s hook frame (text + freeze)
   ↓ 6. append 2s end card (handle + CTA)
   ↓ 7. encode H.264 + AAC, ≤ 100 MB
output.mp4

Reference FFmpeg command (single pass)

ffmpeg -i input.mp4 -i bgm.mp3 -i hook.png -i end.png \
  -filter_complex "\
    [0:v]crop=ih*9/16:ih,scale=1080:1920,setsar=1[main]; \
    [2:v]scale=1080:1920[hook]; [3:v]scale=1080:1920[end]; \
    [hook][main][end]concat=n=3:v=1:a=0[v]; \
    [0:a]volume=1.0[a0]; [1:a]volume=0.25[a1]; \
    [a0][a1]amix=inputs=2:duration=first[a]; \
    [v]subtitles=captions.ass[vout]" \
  -map "[vout]" -map "[a]" \
  -c:v libx264 -preset medium -crf 21 \
  -c:a aac -b:a 128k -ar 44100 \
  -movflags +faststart \
  -t 60 output.mp4

Captions style (.ass)

  • Font: Inter Bold, size 64
  • Outline: 4px black, no shadow
  • Position: lower-third (PlayResY=1920, MarginV=240)
  • Max 2 lines, 18 chars/line, 1.2s min display

Hook frame design

The first 1.5 seconds decide retention. Keep these rules:

  • One big text claim, max 6 words ("I lost ₹50k. Here's why.")
  • Recognisable still from the body, not a generic title card
  • High contrast: dark frame + bright text, or vice versa

What NOT to do

  • Don't centre-crop a horizontal gameplay clip blindly — kill-feed and minimap get lost. Crop with intent.
  • Don't mix BGM louder than -15 LUFS — voice gets buried.
  • Don't burn captions on the bottom 10% — IG UI covers it.

Niche tuning (thenoobstack)

  • Gameplay: hook = score/kill count visible
  • Tech: hook = before/after split-screen
  • AI: hook = one-line claim + tool name
  • Earning: hook = specific number (₹/$), not "make money fast"

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

Check your setup, free