plugins/ffmpeg-social-video/skills/viral-video-hook-templates/SKILL.md
10 proven viral video hook templates with FFmpeg implementations for maximum first-3-second retention. PROACTIVELY activate for: (1) Video hook creation, (2) Opening sequence optimization, (3) Attention-grabbing techniques, (4) Pattern interrupt effects, (5) Curiosity gap implementation, (6) Text overlay hooks, (7) Visual hook effects, (8) Audio hook strategies, (9) Retention optimization, (10) Scroll-stopping content. Provides: Research-backed hook formulas, FFmpeg filter implementations, psychological triggers, platform-specific hooks, A/B testing workflows, and real-world examples with measurable impact.
npx skillsauth add JosiahSiegel/claude-plugin-marketplace viral-video-hook-templatesInstall this skill globally with one command. Works with Claude Code, Cursor, and Windsurf.
3 of 9 scanners reported clean
Some scanners were skipped, did not run, or reported a non-clean status. Review each row below.
MANDATORY: Always Use Backslashes on Windows for File Paths
When using Edit or Write tools on Windows, you MUST use backslashes (\) in file paths, NOT forward slashes (/).
NEVER create new documentation files unless explicitly requested by the user.
The Critical Window: You have 1.3-3 seconds to stop the scroll. Research shows:
| Hook Type | Duration | Reasoning | |-----------|----------|-----------| | Pattern Interrupt (visual) | 0.3-0.5s | Quick shock, then content | | Pattern Interrupt (text) | 1.0-1.5s | Time to read "STOP" etc | | Curiosity Gap | 2.0-3.0s | Two-part reveal timing | | Direct Challenge | 1.5-2.5s | Read + process + react | | Transformation Tease | 2.0-3.0s | Before/after comparison | | Question Hook | 1.5-2.0s | Read + mental engagement |
All FFmpeg filter expressions use SECONDS with decimals:
| Expression | Meaning | Example Use |
|------------|---------|-------------|
| t | Current time | enable='lt(t,2)' = first 2 seconds |
| between(t,0,2.5) | Time range | Hook visible for 2.5s |
| if(lt(t,1),expr1,expr2) | Conditional | Different effect in first second |
| sin(t*10) | Oscillation | 10 cycles per second |
| exp(-t*3) | Decay | Quick fade (3 = fast decay) |
# Hook window timing breakdown:
# 0.0-0.3s: Visual impact (flash, zoom, color)
# 0.3-1.5s: Primary hook text appears
# 1.0-2.5s: Secondary text/elaboration
# 2.5-3.0s: Transition to main content
# Example: Complete hook with timed elements
ffmpeg -i input.mp4 \
-vf "
eq=brightness='0.2*between(t,0,0.2)',
drawtext=text='STOP':fontsize=80:fontcolor=red:x=(w-tw)/2:y=h*0.12:enable='between(t,0.2,1.5)',
drawtext=text='This changes everything':fontsize=48:fontcolor=white:x=(w-tw)/2:y=h*0.20:enable='between(t,1.0,2.8)'
" \
output_timed_hook.mp4
A compact index of the ten canonical viral-video hooks. Each entry has its full FFmpeg filter graph and parameters in references/ten-hook-templates.md.
Load references/ten-hook-templates.md for the full filter graph of any template.
# Subtle shake for emphasis
ffmpeg -i input.mp4 \
-vf "crop=1060:1900:(10+10*sin(t*30)):(10+10*cos(t*25)):enable='lt(t,1)',scale=1080:1920" \
-c:v libx264 -preset fast -crf 23 \
-c:a copy \
output_shake.mp4
# Rhythmic zoom pulse - 8% amplitude at ~2Hz for 1.5s (60% more visible than 5%)
# sin(t*12) = 12 rad/s = 1.91 Hz oscillation. Completes before 1.3-second decision point
# Note: 60fps ensures smooth motion; d=1 for continuous per-frame processing
ffmpeg -i input.mp4 \
-vf "fps=60,zoompan=z='if(lt(t,1.5),1+0.08*sin(t*12),1)':d=1:x='iw/2-(iw/zoom/2)':y='ih/2-(ih/zoom/2)':s=1080x1920" \
-c:v libx264 -preset fast -crf 23 \
-c:a copy \
output_zoom_pulse.mp4
# Red flash then normal
ffmpeg -i input.mp4 \
-vf "colorbalance=rs=0.5:gs=-0.2:bs=-0.2:enable='lt(t,0.3)',colorbalance=rs=0.3:enable='between(t,0.3,0.5)'" \
-c:v libx264 -preset fast -crf 23 \
-c:a copy \
output_color_flash.mp4
# Dark vignette that opens up
ffmpeg -i input.mp4 \
-vf "vignette=PI*min(t,2)/2" \
-c:v libx264 -preset fast -crf 23 \
-c:a copy \
output_vignette_reveal.mp4
#!/bin/bash
# generate_hook_variants.sh - Create multiple hook versions for testing
INPUT="$1"
BASE=$(basename "$INPUT" .mp4)
# Variant A: Curiosity Gap
ffmpeg -i "$INPUT" \
-vf "drawtext=text='Wait for it...':fontsize=64:fontcolor=yellow:borderw=4:bordercolor=black:x=(w-tw)/2:y=h*0.15:enable='between(t,0,2)'" \
-c:v libx264 -preset fast -crf 23 -c:a copy \
"${BASE}_hook_A.mp4" &
# Variant B: Direct Challenge
ffmpeg -i "$INPUT" \
-vf "drawtext=text='STOP SCROLLING':fontsize=72:fontcolor=red:borderw=5:bordercolor=white:x=(w-tw)/2:y=h*0.12:enable='between(t,0,2)'" \
-c:v libx264 -preset fast -crf 23 -c:a copy \
"${BASE}_hook_B.mp4" &
# Variant C: Question Hook
ffmpeg -i "$INPUT" \
-vf "drawtext=text='Did you know?':fontsize=56:fontcolor=white:borderw=4:bordercolor=black:x=(w-tw)/2:y=h*0.15:enable='between(t,0,2)'" \
-c:v libx264 -preset fast -crf 23 -c:a copy \
"${BASE}_hook_C.mp4" &
wait
echo "Generated 3 hook variants for A/B testing"
ls -la "${BASE}_hook_*.mp4"
| Hook Type | Avg Stop Rate | Best Content Type | Risk Level | |-----------|---------------|-------------------|------------| | Pattern Interrupt | 85% | All | Low | | Curiosity Gap | 78% | Educational | Low | | Direct Challenge | 75% | Opinion/Niche | Medium | | Transformation | 80% | Tutorial/Demo | Low | | Bold Claim | 72% | How-to | Medium | | Counter-Intuitive | 70% | Educational | Medium | | Social Proof | 68% | Reviews | Low | | Urgency | 65% | Trending/News | High | | Storytelling | 82% | Personal/Brand | Low | | Question | 74% | Educational | Low |
| Platform | Best Hooks | Avoid | |----------|-----------|-------| | TikTok | Pattern Interrupt, Curiosity Gap, Direct Challenge | Long text hooks | | YouTube Shorts | Transformation, Question, Storytelling | Clickbait-y urgency | | Instagram Reels | Social Proof, Transformation, Bold Claim | Too promotional | | Facebook Reels | Storytelling, Social Proof, Question | Youth slang |
Concrete parameter values for animation timing, color grading LUTs, and audio hook design live in references/animation-color-audio-parameters.md. Highlights:
viral-video-platform-specs - Platform requirementsviral-video-animated-captions - Caption stylingffmpeg-viral-tiktok - TikTok optimizationffmpeg-viral-shorts - YouTube Shorts optimizationdevelopment
This skill should be used when the user asks to train, debug, scale, or improve ML models. PROACTIVELY activate for: (1) PyTorch, TensorFlow/Keras, JAX, Flax, Hugging Face Trainer/Accelerate training loops, (2) distributed training, DDP/FSDP/DeepSpeed, TPU/GPU setup, (3) mixed precision AMP/bf16, gradient accumulation, checkpointing, seeding, (4) overfitting, imbalance, loss functions, regularization, LR schedules, warmup, (5) memory optimization, gradient checkpointing, offloading, quantization-aware training. Provides: reproducible training best practices across deep learning and classical ML.
development
This skill should be used when the user asks to productionize, track, version, govern, monitor, or automate ML systems. PROACTIVELY activate for: (1) MLflow, Weights & Biases, Neptune, Comet, ClearML experiment tracking, (2) model registry, model versioning, artifact lineage, reproducibility, (3) Kubeflow, SageMaker Pipelines, Vertex AI Pipelines, Azure ML pipelines, Databricks workflows, (4) CI/CD, continuous training/evaluation, A/B tests, canary/shadow deployments, (5) drift detection, model monitoring, data validation, responsible AI governance. Provides: end-to-end MLOps architecture and operational safeguards.
development
This skill should be used when the user asks to optimize, export, serve, compress, or accelerate ML inference. PROACTIVELY activate for: (1) latency, throughput, p95/p99, batching, concurrency, KV cache, memory, or cost issues, (2) quantization INT8/INT4, GPTQ, AWQ, bitsandbytes, pruning, sparsity, distillation, (3) ONNX export, ONNX Runtime, TensorRT, TorchScript, torch.compile, XLA, OpenVINO, Core ML, TFLite, (4) Triton, TorchServe, TF Serving, BentoML, Seldon, KServe configuration, (5) edge deployment, CPU/GPU/TPU/Inferentia serving. Provides: hardware-aware inference optimization and safe benchmarking.
testing
This skill should be used when the user asks to tune hyperparameters, run sweeps, optimize search spaces, or use AutoML. PROACTIVELY activate for: (1) Optuna, Ray Tune, FLAML, AutoGluon, Hyperopt, Nevergrad, KerasTuner, W&B sweeps, (2) grid search, random search, Bayesian optimization, TPE, Gaussian processes, evolutionary search, (3) ASHA, Hyperband, successive halving, multi-fidelity optimization, population-based training, (4) learning-rate finder, batch-size search, early stopping, pruning, (5) reproducible sweep design and experiment analysis. Provides: budget-aware hyperparameter search strategy.