Create sourced explainer videos with anything2explainer

Use anything2explainer when an agent needs to turn a topic, article, or document into a finished motion-graphics explainer video instead of handing the user a generic video-generation prompt.

The workflow gives the agent a repeatable production lane: research the topic with source URLs, write narration, build a frame-accurate TTS timeline, storyboard every shot, scaffold the Remotion project, dispatch shot groups to coding agents, render a preview, run frame/QC checks, and deliver the final MP4 with the paper trail.

Invoke it for educational or technical explainers where facts, subtitles, timing, and editable source files matter. Do not treat it as a general video model, presenter-avatar product, Remotion framework listing, or stock-footage generator. The scope boundary is the documented skill process around one deterministic Remotion explainer style: sourced research, narration sign-off, voiceover timing, shot implementation, render, and QC.

Typical operator flow:

1. Clone the repository and link it into the agent skill directory.
2. Ask Claude Code, Codex, or another SKILL.md-capable coding agent to make an explainer video for a topic.
3. Confirm length and language before scripting.
4. Review narration before voiceover.
5. Choose or provide the TTS/audio path.
6. Review the first 30-second pilot render before the full build.
7. Let the agent render, run metrics, fix flagged frames, and deliver the final video plus research/storyboard/QC artifacts.

The upstream project includes a full Remotion template, style references, build scripts, QC rules, prompt templates, and an example film. It supports Chinese and English workflows, defaults to edge-tts for Chinese and kokoro for English, and can accept user-supplied audio when needed.