Video Shortform Skill

Short-form (≤ 10s) is the sweet spot for current text-to-video models — they're great at one shot with one idea, weaker at multi-cut narratives. Plan one shot per call.

Special case: hyperframes-html is not a photoreal text-to-video model. It's a local HTML-to-MP4 renderer. For that model, do not roleplay cinematography or "real-world" camera physics. Treat the brief as a motion design card / title-frame / product interstitial, ask at most one clarifying question, then dispatch immediately.

Resource map

video-shortform/
├── SKILL.md
└── example.html

Workflow

Step 0 — Read the project metadata

videoModel, videoLength (seconds), videoAspect. These are hard-locks — clamp the prompt to whatever the chosen model supports (Seedance 2 caps at 10s; Kling 4 supports up to 10s + image-to-video; Veo 3 supports 8s with audio).

Step 1 — Plan the shot

Write the shotlist BEFORE calling the model:

Slot	Content
Subject	What's in frame?
Camera	Static / pan / push-in / orbit?
Lighting	Key direction + temperature
Motion	What moves, at what pace? Subject motion vs camera motion.
Sound	Ambient bed? (only if the model supports audio)

Normally, show this to the user as a one-sentence plan before dispatching — they can redirect cheaply.

For hyperframes-html, skip the extra pre-dispatch narration once the user has answered the discovery form. Collapse the plan into the actual generation prompt and dispatch immediately.

Step 2 — Compose the prompt

Use the format the upstream model prefers (Seedance: motion + camera + mood; Kling: subject + camera + style; Veo: subject + cinematography + sound). Bind the project's videoAspect and videoLength directly to the API parameters; never put them in prose.

For hyperframes-html, write a concise motion-design brief instead of a camera-realism prompt. Focus on subject, layout, palette, motion character, and overall tone. Do not spend turns narrating environment checks, missing side files, or "I am about to dispatch" status updates.

Step 3 — Dispatch via the media contract

Use the unified dispatcher — do not call provider APIs by hand:

"$OD_NODE_BIN" "$OD_BIN" media generate \
  --project "$OD_PROJECT_ID" \
  --surface video \
  --model "<videoModel from metadata>" \
  --aspect "<videoAspect from metadata>" \
  --length <videoLength seconds> \
  --output "<short-slug>-<seconds>s.mp4" \
  --prompt "<assembled shot prompt from Step 2>"

The command prints one line of JSON: {"file": {"name": "...", ...}}. The bytes land in the project; the FileViewer plays it automatically.

Step 4 — Hand off

Reply with: shot summary, the filename returned by the dispatcher, and one sentence on what to try if the user wants a variation.

For hyperframes-html, keep the reply especially short: what was rendered, the filename, and one concrete variation idea.

Hard rules

• One shot per turn. Multi-shot timelines belong in a hyperframes / interactive-video skill, not here.
• Match videoAspect exactly — re-renders are slow.
• Never ship a video without saving the file — the user expects something to play in the file viewer.
• When the underlying model fails (NSFW filter, content policy, timeout), keep the raw provider error in the tool trace and give the user the media contract's nextStep sentence for that outcome. Never paste upstream error text into the reply. Don't silently retry.
• Do not claim a render has been "sent", "started", or "is running" unless you have already called "$OD_NODE_BIN" "$OD_BIN" media generate.