# video

> A Claude Code skill that produces marketing video using AI generation models, AI avatars, and code-based video frameworks like Remotion — for demos, explainers, social clips, and ads.

**Use case**: Produce video without a production crew

**Canonical URL**: https://agentcookbooks.com/skills/video/

**Topics**: claude-code, skills, marketing, content

**Trigger phrases**: "video production", "AI video", "Remotion", "talking head video", "make me a video"

**Source**: [Corey Haines](https://github.com/coreyhaines31/marketingskills/tree/main/skills/video)

**License**: MIT

---

## What it does

`video` is a Claude Code skill from Corey Haines's [marketing-skills repo](https://github.com/coreyhaines31/marketingskills). It turns Claude into a video producer who picks the right tool per use-case — Veo / Kling / Runway / Sora for AI-generated B-roll, HeyGen / Synthesia for avatar talking-heads, Remotion or Hyperframes for templated programmatic video, ElevenLabs / Cartesia for voiceover. The skill activates when you mention "video production", "AI video", "Remotion", "talking head", or "make me a video".

The output of a session is a video plan: tool selection per segment (avatar, B-roll, motion graphics, voiceover), prompt structures for each AI generation step, Remotion template scope if you're going programmatic, and a cost/time estimate before you spend hours rendering.

## When to use it

Reach for it when:

- You need a product demo or explainer and don't have video resources
- You're producing social clips at scale and want a templated pipeline (Remotion)
- You need talking-head avatar video for an explainer without filming yourself

When *not* to reach for it:

- The video has to feature a real human's actual face for trust reasons
- You're hand-editing a single hero piece; pro tools beat AI for premium one-offs

## Install

The skill is distributed via Corey Haines's [marketing-skills repo](https://github.com/coreyhaines31/marketingskills). Install via the repo's recommended path — copy the [`video` SKILL.md](https://github.com/coreyhaines31/marketingskills/tree/main/skills/video) into your project's `.claude/skills/video/` directory, or use the repo's plugin install if you've set it up.

Once installed, the skill activates on the trigger phrases above. The first time it runs, it will check for `.agents/product-marketing-context.md` (or `.claude/product-marketing-context.md`) — populating that file with your product context first dramatically improves output quality across all of Haines's marketing skills.

## What a session looks like

A typical session has three phases:

1. **Use-case + tool fit.** Demo / explainer / social clip / ad / programmatic? Each maps to a different toolchain. The skill picks instead of listing five options.
2. **Pipeline design.** For programmatic: Remotion template scope, data input, render cost estimate. For AI-generated: prompt structures and tool sequencing (script → voiceover → avatar → B-roll → composition).
3. **Cost + time check.** Rendering AI video at scale gets expensive fast; the skill estimates before you start so the budget is real.

The discipline that makes it work: matching tool to use-case. Asking Veo to render a templated thumbnail is wasteful; asking Remotion to render lifelike humans is futile. Tool fit is the entire game.

## Receipts

Honest reporting on what `video` produces and where it has limits:

**Where it works well:**
- The tool-fit decision saves the "I tried Runway, Kling, and HeyGen and got mediocre results in all three" loop
- Cost estimate up front is genuinely useful — AI video burns budget invisibly
- Remotion-pipeline scope for templated video at scale is the highest-leverage pattern

**Where it backfires:**
- AI video quality shifts every few months; prompt structures need ongoing tuning
- Avatar-tool talking heads are good but still readable as AI to careful viewers; brand-trust use-cases need real video

**Pattern that works:** use AI video for repurposing and scale (variations, social clips, explainers), reserve real video for the moments that need a face. The split keeps both volume and trust high.

## Source and attribution

Originally written by [Corey Haines](https://corey.co). The canonical SKILL.md and any supporting files live in the [`video` folder](https://github.com/coreyhaines31/marketingskills/tree/main/skills/video) of his [marketing-skills repository](https://github.com/coreyhaines31/marketingskills).

License: MIT. You can install, adapt, and redistribute the skill, with attribution preserved.

This page documents the skill from a practitioner's perspective. For the formal spec and any updates, defer to the source repo.