Stickman Video Director: Turn Plain Text into a Fully Directed 60-Second Animation
Most AI video tools hand you random clips. You type a prompt, you get ten seconds of a character standing in a room, and it’s on you to stitch something coherent out of the noise. Stickman Video Director takes a different approach: it acts as a professional director, breaking your idea into six distinct scenes with specific camera angles and visual metaphors — before you burn a single generation credit.
It’s a Codex Skill (MIT licensed), which means there’s no API to wire up and no MCP server to run. You install the skill, invoke it in conversation, and it walks the whole production through with you.
A script is not yet a video
The core insight is a good one. A solid idea can still produce a flat animation: one character, one background, ten seconds with nothing new to look at. Directing a full minute is a different craft — you have to shape the opening hook, pace the explanation, invent relevant visual metaphors, move the camera, connect scenes, and protect visual continuity across separate generations that don’t know about each other.
Stickman Video Director does that production thinking up front. The result is a storyboard you actually approve before anything renders.
How it works: paste → choose → confirm → generate → stitch
- Paste your copy, notes, an article — or just give it a topic.
- Choose an aspect ratio (
16:9,9:16, or1:1) and a light or dark theme. - Confirm a detailed six-scene director’s proposal — complete with English voiceover, a reference translation, shot descriptions, camera moves, transitions, background music, and sound effects.
- Generate six standalone Gemini Omni Flash prompts once the proposal is approved.
- Stitch the six ~10-second clips into a coherent one-minute video.
The approval checkpoint is the part that matters. You revise the story while changes are still cheap — at the storyboard stage — rather than after you’ve spent credits generating scenes that don’t hang together.
Why the output holds up
A few design choices push this beyond “prompt and pray”:
- Story architecture — raw material gets reworked into a sharp opening, progressive explanation, and a closing callback, while preserving the central meaning.
- Dense, relevant motion — three timed beats per clip, each with visual metaphors, environment changes, camera movement, on-screen text, interactions, and transitions.
- Production locks — every standalone prompt repeats the character, line weight, palette, voice, dialogue, audio, transition, and negative constraints. That’s how six independent generations stay visually consistent.
- Format-aware direction — staging, camera paths, and text placement are recomposed for each aspect ratio rather than just swapping a ratio label.
- Faithful to source — it won’t invent facts, data, quotes, or product claims that aren’t grounded in your input.
The final deliverable is a full package: a creator-facing title and hook, roughly 130–150 words of one-minute narration, six distinct scenes, precise English dialogue with reference translations, and six timed Gemini Omni Flash prompts with negative constraints and stitching guidance.
Who it’s for
If you’re making educational explainers, motivational stories, or fast-moving visual essays for YouTube Shorts, TikTok, Instagram Reels, or long-form YouTube, this is a shortcut to a polished, professional result without spending hours on manual animation. The high-contrast stickman aesthetic — black figure on white, or white figure on black, with up to three saturated accent colors — is deliberately simple, which is exactly why it reads well on a phone screen.
The honest caveat: this is a director, not a renderer. It produces the storyboard and the prompts; the actual video comes from Gemini Omni Flash, so your output quality and cost still ride on that generation step. But that’s arguably the right division of labor — the thinking happens where revision is free, and the generation happens once the plan is locked.