Agent Skills

glass-shard-title

videoheygen-com3K installs

Glass shards fly in through fog and tile themselves into the headline. HyperFrames block, 1920×1080, 12.16s, 19 variables.

Install

npx skills add https://github.com/heygen-com/hyperframes --skill glass-shard-title
SKILL.md

Glass Shard Title

Voronoi-cut glass panes fly in from depth, rotating, and land as a tiled silhouette of the headline, then fly out past the camera. Each shard is a Sutherland–Hodgman clip of a jittered blob against its Voronoi cell, so the pieces always cover the type. Glass and matcap surfaces get per-fragment fog; the z-flight is decoupled from the ease so shards visibly traverse it. Tile count 1 gives a single pane.

Composition id: glass-shard-title. Duration 12.16 s at 30 fps, 1920×1080.

Files

  • glass-shard-title.html (16 KB)
  • assets/ferndale_studio_01_1k.hdr (1.6 MB)
  • assets/fonts/Geist-Bold.ttf (65 KB)
  • assets/fonts/Geist-Regular.ttf (65 KB)
  • assets/fonts/Geist-SemiBold.ttf (65 KB)
  • assets/fonts/Geist-OFL.txt (4 KB)
  • assets/fonts/cormorant-garamond.woff2 (37 KB)
  • assets/fonts/CormorantGaramond-OFL.txt (4 KB)
  • assets/glass-main.js (549 KB)
  • assets/Three-LICENSE.txt (1 KB)
  • assets/matcap-1.png (fetched from the CDN at install)

Install

Install with npx hyperframes add glass-shard-title; by default the files above land under compositions/glass-shard-title/. Then mount the block from the host index.html:

<div
  data-composition-id="glass-shard-title"
  data-composition-src="compositions/glass-shard-title/glass-shard-title.html"
  data-start="0"
  data-duration="12.16"
  data-track-index="1"
  data-width="1920"
  data-height="1080"
></div>

Render with custom values by targeting the composition file directly:

npx --yes hyperframes@0.8.12 render 'compositions/glass-shard-title/glass-shard-title.html' --variables '{"headline":"Designed in glass","tileCount":8}'

Variables

Read at runtime via window.__hyperframes.getVariables(); declared on the composition root as data-composition-variables (single-quoted attribute, plain JSON).

id type default label / range
headline string "Designed in glass" Headline max 40 chars
tileCount number 8 Tiles 1–400 step 1
roundness number 0.4 Roundness (0-1) 0–1 step 0.01
bevel number 0.35 Bevel (0-1) 0–1 step 0.01
meshSmooth number 0.6 Surface smoothness (0-1) 0–1 step 0.01
flyInTime number 2.3 Fly-in time (s) 0.2–6 step 0.05
flyOutTime number 2.4 Fly-out time (s) 0.2–6 step 0.05
stagger number 0.27 Stagger (0-1) 0–1 step 0.01
flyInRotation number 5 Fly-in rotation (0-60) 0–60 step 0.1
flyOutRotation number 6 Fly-out rotation (0-6) 0–6 step 0.1
chaos number 0.25 Outline chaos (0-1) 0–1 step 0.01
padding number 1 Edge padding (0-1) 0–1 step 0.01
sizeVariance number 0.35 Size variance (0-1) 0–1 step 0.01
gap number 0.03 Tile gap 0–0.3 step 0.002
easePow number 10 Fly-in position ease (1-10) 1–10 step 0.1
sideDist number 3 Side fly distance 0–60 step 0.5
zDist number -42 Z fly-from (negative) -400–0 step 2
zOut number 14 Z fly-out to (+ = past camera) -400–60 step 2
fog number 36 Fog distance (0 = off) 0–400 step 2

Runtime contract

  • One paused GSAP timeline registered as window.__timelines["glass-shard-title"].
  • Re-syncs on the hf-seek CustomEvent; every frame is a closed-form function of time (seeded PRNG only, no rAF loops, no Date.now).
  • Renderer: three.js 0.181.2 (bundled), GSAP 3.14.2, Canvas 2D, GLSL shaders, HDR environment, Matcap, Seeded PRNG, d3-delaunay, Shadow maps. Budget roughly 350 MB per live instance; run one at a time.
  • External runtime dependencies: https://cdn.jsdelivr.net/npm/gsap@3.14.2/dist/gsap.min.js, https://cdn.jsdelivr.net/npm/d3-delaunay@6.0.4/dist/d3-delaunay.min.js.
  • Local fonts: assets/fonts/Geist-Regular.ttf, assets/fonts/Geist-SemiBold.ttf, assets/fonts/Geist-Bold.ttf.
  • Bundled code is inlined into the composition (assets/glass-main.js); rebuild with esbuild and re-inline after editing the .mjs source.

Editing rules (from the source project)

  1. Keep data-composition-variables a single-quoted attribute with plain " JSON. Never save it through Studio's Design panel.
  2. Do not put <canvas> in static markup; create it at runtime.
  3. Keep every visual state a function of t; seek-safety is what makes the block renderable.

Related skills

video-editgenmedia-labs715KEdit existing video on RunComfy — this skill is a smart router that matches the user's intent to the right edit model in the RunComfy catalog. Picks Wan 2.7 Edit-Video (general restyle / background swap / packaging swap, identity + motion preservation), Kling 2.6 Pro Motion Control (transfer precise motion from a reference video to a target character), or Lucy Edit Restyle (lightweight identity-stable restyle / outfit swap). Bundles each model's documented prompting patterns so the skill gets shai-video-generationgenmedia-labs714KGenerate AI videos on RunComfy via the `runcomfy` CLI — a smart router across the full video-model catalog: HappyHorse 1.0 (Arena #1, native in-pass audio), Wan-AI Wan 2-7 (open weights, audio-driven lip-sync), ByteDance Seedance v2 / 1-5 / 1-0 (multi-modal cinematic), Kling 3.0 / 2-6, Google Veo 3-1, MiniMax Hailuo 2-3, ByteDance Dreamina 3-0. Covers text-to-video (t2v), image-to-video (i2v), and Veo's video-extend endpoint. The skill picks the right model for the user's intent (Arena-#1 qualitai-musicgenmedia-labs714KGenerate AI music on RunComfy via the `runcomfy` CLI — a smart router across the music-model catalog. Routes to ElevenLabs AI Music Generation (premium 44.1 kHz stereo vocal tracks, 5 s–5 min, $0.0083/s) and ACE Step / ACE Step 1.5 (StepFun-AI open-weights, tag-driven composition, multilingual lyrics, $0.0002–0.0003/s, ~27× cheaper), plus ACE Step audio-inpaint (regenerate a time range inside an existing track) and ACE Step audio-outpaint (extend a track before or after). Picks the right model fimage-to-videogenmedia-labs713KAnimate any still image on RunComfy — this skill is a smart router that matches the user's intent to the right i2v model in the RunComfy catalog. Picks HappyHorse 1.0 I2V (Arena #1, native audio, identity preservation) for general animations, Wan 2.7 with `audio_url` for custom-voiceover lip-sync, or Seedance 2.0 Pro for multi-modal animation from image + reference video + reference audio. Bundles each model's documented prompting patterns so the caller gets sharper output without burning iterat

Search skills and MCP servers

Fuzzy search across 23,137 skills and servers