sota-animated-landing-page
Skill9 starsChanged 11 days agoMITView on GitHub
Research and build art-directed landing pages with reference-matched imagery, expressive typography, cinematic scroll sequences, responsive layouts, and visual verification. Use for landing-page creation or substantial redesign while preserving the existing product and the user's chosen visual direction.
In this repository2 other files it gives its agents.
- readme-showcaseSkill
- stageshowSkill
The file
We split this skill into the rules it gives an agent
10 rules. About 0 are common, 9 are distinctive, and 1 is very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈9
Distinctive · estimate
This author's own choices. Worth a look.
1
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 10
Read the existing app, local agent instructions, task state, and available assets.
- Distinctive: line 11
Use the concept and references to determine the hero: an architectural reveal, interface assembly, character scene, neural core, or another …
- Distinctive: line 13
Treat discovery, downloadability, use permission, and production readiness as separate facts.
- Distinctive: line 14
Follow a supplied font collection to the creator/foundry, select display and interface roles, and confirm actual webfont terms and files.
- Distinctive: line 19
Read the animation production workflow when creating the hero.
- Distinctive: line 27
Keep the headline, explanation, navigation, and primary action semantic HTML outside the canvas.
- Very specific: line 28
Use a 400vh desktop track and a sticky 100vh canvas as the baseline.
- Distinctive: line 33
Avoid loading dozens of videos or full-resolution posters at once.
- Distinctive: line 37
Check desktop and narrow mobile layouts, keyboard navigation, visible focus, contrast, reduced motion, image failures, loading states, and …
- Distinctive: line 39
Run relevant builds and behavior checks.
≈ “Common” is an estimate: at least 1 in 5 of the 372,429 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.
Files
SOTA Animated Landing Page
Brief and evidence
- Read the existing app, local agent instructions, task state, and available assets. Preserve working user journeys and data. If the user requests research and a plan first, produce those before product implementation.
- Use the concept and references to determine the hero: an architectural reveal, interface assembly, character scene, neural core, or another appropriate subject. Never force a commercial product shot.
- Browse actual visual references. Record source URLs, what each contributes, and whether imagery has been visually inspected. Source-led imagery must match the named subject, character, series, and era; don't substitute stock photos or unrelated fan art.
- Treat discovery, downloadability, use permission, and production readiness as separate facts. Record actual status instead of making unsupported availability or license claims.
- Follow a supplied font collection to the creator/foundry, select display and interface roles, and confirm actual webfont terms and files. Keep reading text legible; display fonts belong in large headlines.
- Define composition, palette, type pairing, graphic language, interaction, and mobile behavior. Awwwards-level ambition is a brief, not proof of an award or measured quality.
Production sequence
Read the animation production workflow when creating the hero. It includes baseline settings for Google Flow, Nano Banana Pro, Veo, FFmpeg, and Stitch, along with filenames and scroll behavior.
Check actual tool access before starting. Prefer the requested provider when available. If Flow, Veo, Stitch, or browser integration is unavailable, complete independent research, reference collection, storyboards, layout, and code scaffolding, and report which provider-specific outputs remain unavailable. Do not claim another generator ran the requested service. Spending and publication remain within the current task's authorization.
For established-character imagery, anchor the work to verified source images. A current web-sourced-art requirement takes precedence over generating replacement character art. The generation workflow can animate backgrounds and panel reveals around verified artwork; disclose synthetic character motion and inspect for identity drift.
Assembly
- Keep the headline, explanation, navigation, and primary action semantic HTML outside the canvas. The page must communicate before animation loads and if it fails.
- Use a
400vhdesktop track and a sticky100vhcanvas as the baseline. Map normalized scroll to a frame index, paint viarequestAnimationFrame, and fade hero copy after 30% progress. Preserve native wheel and touch scrolling. - Separate compressed frame prefetch from decoding. A sequence of 210 decoded 1920×1080 RGBA frames requires roughly 1.62 GiB before browser overhead. Start with the poster and a small decoded window. Prefetch the remaining compressed assets after critical content loads, within connection and memory limits. Explain the cost if full decoded preloading is required.
- Provide a static/reduced-motion path and a short mobile layout when the desktop sequence would impede access. Cap rendering resolution, resize correctly, and cancel observers, frames, and media when leaving the view.
- Adapt imported Stitch output to the real component system. Remove duplicated hero/nav content, replace invented content, preserve chosen fonts, and make visible actions work.
- Demonstrate the actual product through a useful preview, choices, benefits, honest FAQ, and a final action. Do not invent customers, testimonials, awards, usage counts, or integrations.
- Avoid loading dozens of videos or full-resolution posters at once. An ambient dashboard loop is separate from a beginning-to-end hero transition; read the loop criteria in the production reference.
Verification and handoff
- Check desktop and narrow mobile layouts, keyboard navigation, visible focus, contrast, reduced motion, image failures, loading states, and the main user journey.
- Inspect rendered screenshots when a browser is available. Report when visual verification has not been possible; a successful build is not visual verification.
- Run relevant builds and behavior checks. Verify data preservation and account isolation when persistence or sign-in changes.
- Measure critical loading and interaction; identify media bottlenecks. Label targets separately from measured results.
- Hand off source files, provenance, dependencies, and verified local results. Push to GitHub or deploy to Vercel when requested or already authorized, following the project's hosting workflow.
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
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.

