Input format guide
What Motion Content Studio reads from your AI director's reply, how to test your own text, and a template you can copy.
How it works
- Get the reply. In the app, the AI Director mode can run the director for you and fill the project as it goes; this guide describes the text it (or you) must produce. Or paste the whole conversation yourself: Paste the whole reply. Copy the full conversation output from your AI director, steps and all, into the Import & Parse tab. Everything it finds shows up in the YouTube (16:9) and Reels (9:16) tabs.
- Press Parse to read it. The text is read in your browser. While you type you only see a preview; nothing in the project changes until you press Parse (or Ctrl/Cmd + Enter). In the demo nothing leaves your browser.
- Check the counts. Chunks, scenes, animations, reels, covers and publishing text are listed with warnings if something looks off.
Sections and labels the parser recognizes
| Part | What to write | Notes |
|---|---|---|
| Step 1: audio chunks | Each chunk starts with the chunk start phrase. Default: | A chunk runs until the next chunk, a heading, ---, a code fence, a Step N line or a label. The phrase is a setting, so it can be anything. |
| Step 2: image scenes | Scene Number: [01], then Narration Part: "...", then Image Prompt: followed by a code block. | Persian and Arabic-Indic digits are accepted in scene numbers. A scene without a prompt gives a warning. |
| Step 4: video prompts | A line naming the scenes to animate, then Scene Number: [02] and Video Prompt: with a code block. | Only scenes named here get an animation. They are not counted as image scenes. |
| Step 5: thumbnail | Thumbnail Prompt: followed by a code block (the 16:9 YouTube cover). | Read from Step 5, or anywhere before Step 6 when there are no step headings. |
| Step 6: reels | A heading Step 6: Reels and Shorts, then Reel 01 to Reel 03. Each reel has an Audio chunk that starts with the start phrase, Reel Scene Number: [R01-01] with an Image Prompt (3 to 4 vertical 9:16 prompts), one or two Video Prompt (Flow 9:16) blocks and a Reel Cover Prompt:. | Everything inside Step 6 stays separate from the main video counts. A missing cover is a warning. |
| Publishing text | Cover Headline:, YouTube Title: and Caption:. Inside a reel block they belong to that reel. | Missing ones are simply left empty. Hashtag lines under a caption are kept. |
What is tolerated and what breaks it
Tolerated
- Bold markdown on labels (
**Scene Number: [01]**or plain). - Persian, Arabic-Indic or Latin digits.
- Code fences with
```or~~~. - ZWNJ and direction marks, and the Arabic forms of yeh and kaf used in place of the Persian letters.
- Label variants such as
Cover Prompt,Reel Caption,Video Title, and a full-width colon. - Director notes between sections.
What breaks it
- Chunks that do not start with the start phrase. The phrase is configurable: set it in Import & Parse.
- Prompts written outside a code block on a different line from their label.
- Reels written before a Step 6 heading, so they count as main scenes.
- The same scene number used twice.
- Fewer than 3 or more than 4 image prompts in a reel gives a warning.
Minimal template
A short skeleton with placeholders in square brackets: 2 chunks, 2 scenes with 1 animation, a thumbnail, publishing text and 1 reel. Replace the brackets with your own text. It uses the default start phrase.
Check my text
Paste your own text to see what the real parser finds. It runs in this page only and creates no project.
Your own format?
The start phrase and the content language are settings, so a different narrator refrain or language works without any change. If your director writes labels differently, the parser rules can be adapted for your channel. Get in touch through pooya.website.