Animated learning video

The person brings what is to be learned (a lesson plan, a topic and an age, a doc or a training page) and who it is for. You deliver a 1–3 min 16:9 video in which one or two recurring guide characters teach it, one idea per scene, narrated, with the key words and diagrams on screen and a recap at the end. Rounds and deliverables follow the core skill.

Steps

1. The learning beats (guide-make-a-script-from-source, a source to retell; the learner's age and what they should be able to do afterwards, from the person). The shape this workflow adds: the question the lesson answers → one idea per scene, each shown before it is named → a recap of the ideas in order → one question that checks understanding, with a pause to answer it. A read slower than an ad (guide-make-a-script-from-source, Length: check the take's words per second). Gate: the beat table (narration, picture, on screen) and the length.
2. The look and the guides (guide-pick-a-stylized-look): style frames of the opening scene in two or three looks (a drawn 2D cartoon, 3D animation, paper-cut), then a sheet per guide character (nano-banana-character-sheet). A guide's name on their cap or shirt is baked lettering, spelled in quotes. Gate: the look, then the guides.
3. The set, when scenes share a place (a playground, a classroom, a kitchen: guide-design-a-set), designed once and staged into (nano-banana-stage-cast-in-set).
4. Stills (guide-pick-a-stylized-look, Process step 4): the props of each idea as objects the guides handle (cones, blocks, a ball), no text in the pictures. Gate: the stills.
5. One motion test, then the batch (guide-animate-a-shot, h3-max-image-to-video): mouths closed, the guides demonstrating each idea with their hands. Gate: the test's motion.
6. Narration (guide-cast-a-voice, then guide-voice-a-script, A narrator for a short): a warm, clear voice for the age; the pick is the take and its alignment gives the word times.
7. Base cut (ffmpeg-cut-clips-to-a-voiceover), a quiet bed if the person wants one (elevenlabs-generate-music).
8. Key words, diagrams and captions (guide-add-titles-and-motion-graphics): each idea's word appears as it is said, diagrams built step by step on the words (hn4a-compositor-motion-graphics), a calm camera on any still diagram (hn4a-compositor-calm-explainer-camera), calm captions on by default for learners (hn4a-compositor-in-scene-titles, captions); the recap lists the ideas as they are named; the check question holds on screen through its pause. Gate: the live spec (hn4a-compositor, Live spec, then render).
9. Deliver per the core skill; a 9:16 cut if it will be posted to social (guide-pace-a-short-for-social).

Choices this workflow makes

• Show, then name. Each idea is acted out by the guides before the narration names it, and the word appears on screen as it is said.
• One idea per scene, so a learner can rewatch a scene for one idea.
• Recurring guides the learner gets to know, on sheets so they stay on model.
• Words on screen only from the compositor, never in the pictures, so they are spelled right and can change for another language.

Dead ends

• None recorded on this workflow yet; the traps it inherits are in guide-pick-a-stylized-look and nano-banana-character-sheet. Record the job's own as it runs.

Worked example

The shape of the demo clip: a preschool movement lesson with two cartoon kangaroo guides (a big one in a cap that reads "ROO", a small one whose cap reads "PIP") on a school playground with coloured cones and a ball, in a bright drawn 2D look. 16:9, 30 fps.

At list prices for a 90 s lesson (two sheets, a set, about 15 stills, about 90 s of H3 Max at 768P, the narration): about $10–12, the cut and graphics free.