SkillMachine
All skills

AI Video Production Producer

Document

From brief to publish-ready video package: the right tool, script, and render pipeline

Live output preview

Input Format: Input FormatOutputWatch the Output: Watch the Output

A plan is required to view this content

Choose a plan to access input format, sample outputs, and live previews.

View Plans →

About the skill

What it does

It takes a video production brief (type, platform, length, presenter, assets, budget) and produces a publish-ready production package: it selects the approach, writes a scene-by-scene script, prepares the AI model prompts, fixes the aspect ratio and render pipeline, and ships it with a production checklist. It works strategy-first: decide what video is needed before picking tools, never the reverse.

It separates four approaches with clear criteria: Programmatic (Hyperframes = LLM-native HTML/CSS, Apache 2.0; Remotion = React, advanced animation + Lambda batch) for templated, data-driven, recurring video; AI Generation (Veo 3, Sora 2, Runway Gen-4, Kling, Seedance, Hailuo) for B-roll and hero shots you can't film; AI Avatars (HeyGen with MCP, Synthesia) for talking-head without filming; Repurposing (Descript, Opus Clip, CapCut, Captions.ai) to cut long-form into clips. Model choice rests on a cost/quality trade-off: Veo/Sora for top quality + synced audio; Kling/Seedance for volume at low cost (~$0.09-0.14/sec); Hailuo for character consistency across shots; Hunyuan/Wan for self-hosting.

When to use

When you want to consolidate every production decision into one package before producing a demo, explainer, social clip, ad, or tutorial — especially when tool overload makes the choice hard (which model? avatar or VO? programmatic or generation?), when turning a template into a batch, or when feeding the video step of an agent pipeline with a machine-readable spec.

Method / frameworks

Subject + Action + Camera + Style + Mood prompt formula (for generation); a 4-approach selection matrix; a model comparison table (resolution × duration × cost × strength); platform → aspect-ratio mapping (9:16 social, 16:9 YouTube/web, 1:1 feed); attention-discipline rules (3-second hook, captions mandatory — ~85% of social video is watched silent); and an agent-native pipeline (script → Hyperframes/HeyGen MCP/model API → assembly → publish). Output is dual: output.json (agent-readable) + output.html (marketplace visual).

How do I use this skill?

You don't "run" a skill — after installing it you just tell the agent your task (e.g. ask for the relevant job), and the skill kicks in by itself when its description matches.

Upload the video.zip you downloaded as-is — no packaging needed, the format is already correct (folder at root).

  1. Open Settings → Customize → Skills
  2. Upload → select the video.zip you downloaded
  3. Claude reads SKILL.md; the name + description appear. Ready ✅

Scripts run in Anthropic's code-execution environment (sandbox) — not on your machine.