47 comments

[ 5.2 ms ] story [ 48.8 ms ] thread
Have you thought about integrating Remotion into this?
What do you think the next update will be about..^_~
Just watch out for their license...
Oh!! yeah that can be an obstacle.. no issues, I guess I found Mediabunny at the right time..
A demo video would be good.
Yeah actually, that's a great idea, working on it ..(~ ̄▽ ̄)~
Yeah I'm sure you need fable for the heavy reasoning task of editing tiktoks
My impression was that they made this editor with Fable, and its JSON project structure would only serve well for manipulation by lesser models.
Very cool project but having fable in the name makes it sound so tacky
The readme is full of compound Claudeisms, so why not ;-)

    per-shot energy

    frame-accurate

    word-pop

    reference-video

    CSS-@keyframes-animated SVGs render frame-accurately
That’s just how some humans write. These words didn’t come from an LLM vacuum.
Ya, a demo video would be nice .
I have no use for this, however I am stunned by the simple and effective solution of using json+sse instead of an API to allow agent interaction Amazing work
I have little experience with video editing; a tool like this might be an ideal path to learn more. Thanks for creating and publishing it and sharing here!
You aren't going to learn by having the AI do it for you...
Agreed. I didn't mean to imply otherwise. It was more about having every facet represented in JSON in one place.
This is pretty dang cool but I guess I was hoping to see some very clever use of vision and maybe waveform analysis to give the model "eyes and ears" to edit audio and video. V2
Cool concept, will try it out! I've had decent results with computer use operating conventional editing tools, but being able to directly edit JSON project files is a solid optimization and opens up a lot of opportunities with things like modular templating.
you can just give claude code access to ffmpeg
You should have a hosted site for this - It is entirely frontend - right?
You should strongly consider using Mediabunny to do all of the rendering client-side!
[flagged]
Can you please not post AI-generated or AI-edited comments to HN? It's not allowed here - see https://news.ycombinator.com/newsguidelines.html#generated and https://news.ycombinator.com/item?id=47340079.

Of course, it's impossible to know for sure what was LLM processed or not, but some of your posts (like this one) have been getting classified that way.

I apologise for the activity as some of my replies were AI-assisted. As I wanted to be precise and frame my idea accordingly. I'll keep the guidelines in mind and maintain the legitimacy of my comments and post.
PSA stop putting bigco IP in your project names!

has clawdbot taught us nothing?

Guess it is efficient for attention.
This looks great, will try it out. I had a crazy idea of doing a full agentic demo video narration and cutting, so might try your tool with some tweaks.
Thanks! That sounds like a really interesting use case. I'd love to see how it turns out. If you end up making those tweaks for full agentic narration and video cutting, feel free to share your experience or ideas, I'd be happy to learn from them and see if they could be incorporated into the tool.
Love that the timeline is just JSON and the UI hot-reloads. Agents and humans on the same edit is the interesting part.
demo is up, as promised: https://github.com/ronak-create/FableCut/releases/download/v...

its a real session, not staged: one prompt ("edit a reel from these assets"), claude code plans the cut for ~6 minutes, writes the entire timeline as one json document, and the editor is already live with it. the 24.5s reel it produced is also in the release, untouched: https://github.com/ronak-create/FableCut/releases/tag/v1.3.0

fittingly, the demo video itself was also edited in fablecut.

Super cool, but maybe it is worth considering using OTIO format instead of the simple JSON for a timeline? LLMs are pretty good for handling OTIO
Nice idea, let me give it a go-through. Thanks btw!!
I've looked at it and i've found thatit is more verbos for LLMs not less, curently the format is optimized for cheap agent patching, and OTIO support as an interchange adapter (not a native format swap).