Plan videos with shot lists, scripts, storyboards, and filming guides for social media.
Shot-by-shot storyboards for TikTok, Reels, YouTube Shorts, and UGC ad scripts. Built around retention physics, not film-school conventions.
Activate this skill when the user says any of the following (or similar phrasing):
"storyboard," "shot list," "shot-by-shot breakdown"
"video script," "script for my video," "TikTok script," "Reel script"
"creator brief," "filming guide," "filming plan"
"UGC ad script," "UGC brief"
"pre-production plan"
"hook ideas," "hook variants," "video hooks"
"plan my next video," "plan a video"
"video ad script"
"content series for video," "plan a video series"
"plan a TikTok / Reel / Short" — planning what to film, not auto-generating a rendered video
"video content plan" — planning a filming schedule, not writing blog posts
"social media video ideas" — brainstorming video concepts, not written captions
"create a TikTok ad" — when the user intends to film it themselves (ask if ambiguous)
"ad campaign for [product]" — when the user mentions video, filming, or creators (otherwise may be ad-creative for static ads)
If the user says "make me a video," "create a TikTok," or "video ad" without specifying whether they want a plan to film or a rendered animation, ask:
"Would you like me to plan a video for you to film (storyboard with shot list and script), or create an animated video that renders automatically?"
Static ad images → use ad-creative skill instead
Written social media posts, blog content, email copy → use content-machine skill instead
Brand identity (logos, colors, typography) → use branding-generator skill instead
Code-rendered animated videos (motion graphics, animated explainers that play automatically) → use video-js skill instead
Slide decks / presentations → use slides skill instead
SEO landing pages → use programmatic-seo skill instead
| This skill (storyboard) | video-js | ad-creative | content-machine |
|---|---|---|---|
| Plans videos you go film yourself|Rendersanimated videos from code | Createsstatic ad images| Writestext content (posts, blogs, emails) |
| Output: script + shot list + visual storyboard on canvas + creator brief | Output: playable animated video in browser | Output: designed image files | Output: written copy |
| For real-camera, UGC, creator content | For motion graphics, explainers, animated ads | For display ads, social image posts | For captions, articles, newsletters |
| Platform | Ratio | Resolution | Max duration | Sweet spot | File |
|---|---|---|---|---|---|
| TikTok | 9:16 | 1080×1920 | 10 min in-app / 60 min upload | 21-34 sec | MP4/MOV, H.264, 287MB mobile / 500MB web |
| Instagram Reels | 9:16 | 1080×1920 | 3 min in-app / 15 min upload | <90 sec for Explore boost | MP4/MOV, 4GB |
| YouTube Shorts | 9:16 | 1080×1920 (up to 4K) | 3 min | 15-60 sec | MP4 |
| YouTube long-form | 16:9 | 1920×1080+ | unlimited | 8-12 min (mid-roll ads) | — |
Universal safe zone for cross-posting:Keep all text, faces, logos inside thecenter 900×1400 of the 1080×1920 frame. Top 14% + bottom 20-35% are covered by UI on at least one platform.
TikTok's algorithm scores hook retention separately from total watch time. 2025 creator analytics:
| 3-sec retention | View multiplier | Outcome |
|---|---|---|
| 85%+ | 2.8× | Viral tier — FYP push |
| 70-85% | 2.2× | Optimal reach |
| 60-70% | 1.6× | Average |
| <60% | baseline | Minimal distribution |
Target: keep ≥65% of viewers past 0:03. If you're losing >35% in 3 seconds, the hook is broken — rewrite the opening, not the body. 84% of viral TikToks in 2025 used an identifiable psychological trigger in the first 3 seconds.
The scroll-stopping element must fire in 0-2 seconds. Seconds 3-5 expand it. Never introduce — interrupt. Banned openers: "Hey guys," "Welcome back," "So today I'm gonna..."
| Hook | Template | Trigger |
|---|---|---|
| POV | "POV: you just found out [revelation]" | Puts viewer inside the scenario; personal relevance |
| Stop-scrolling callout | "Stop scrolling if you're a [role] who [pain]" | Audience self-selects; filters for high-intent |
| Contrarian | "Everyone says X. That's completely wrong." | Cognitive dissonance demands resolution |
| Unfinished story | "I almost [drastic action] until I found..." | Open loop — Zeigarnik effect |
| Negative listicle | "3 [category] mistakes that are costing you [outcome]" | Loss aversion > gain framing |
| Number hook | "$47,000 in 30 days — here's the exact breakdown" | Specificity = credibility |
| Secret reveal | "What [authority] doesn't want you to know about X" | Insider info promise |
| Surprise reaction | Open on a shocked face, silent beat, then reveal | Viewer's brain asks "what are they reacting to?" |
| Visual interrupt | Start mid-action, mid-motion, mid-chaos | Pattern break — no static frame 1 |
The silent test: Watch your first 3 seconds on mute. If text overlay + visual alone don't communicate the promise, it fails — ~85% of social video is watched muted.