One essay in. Every platform's version of it, out.

Ingest

Get the essay into the studio

Two ways in, both automatic once set up.

Substack RSS feed

Point the studio at your Substack's RSS feed (the automatic, machine-readable feed every Substack publishes of its own posts) and new essays are picked up on their own as you publish them. No manual export step.

Paste an essay

Don't use Substack, or writing something that isn't on it yet? Paste the essay text in directly and it runs through the same pipeline.

Obsidian vault import

If you draft in Obsidian (a local, folder-of-text-files note-taking app), the studio can import essays straight from your vault, so the source of truth stays wherever you already write.

Generate

Turn one essay into everything else

A single run produces a full set of platform-native drafts, not one generic blurb copy-pasted everywhere.

LinkedIn post + first comment

A LinkedIn-shaped post, plus a first comment drafted to go with it (a common way to add a link or a follow-up thought without cluttering the main post).

X / Twitter thread and a standalone tweet

Both a multi-post thread version of the essay's argument and a single stand-alone tweet, generated separately.

Instagram caption, Threads post, Bluesky post

A caption written for Instagram, and native posts for Threads and Bluesky, each respecting that platform's tone and length.

6 to 9 slide carousel, rendered as real PNGs

Not a template you fill in by hand: the studio outputs actual image files, ready to upload as a carousel post.

Quote cards

Standalone image cards pulling the strongest lines out of the essay, for posting on their own.

Faceless vertical video

An MP4 with burned-in captions (text rendered directly into the video frames, so it's readable with the sound off) in 9:16, 1:1 or 16:9, with no camera or face required.

Narrated audio essay

The essay read aloud as a standalone audio file.

Voice

Narrated in your own speaking voice

Writing voice and speaking voice are cloned separately, and the speaking side is where it gets literal.

One recording, once

Record about a minute of yourself speaking, one time. That sample is what reels and audio essays are narrated in going forward.

Generated locally, on your own GPU

Narration runs on your own machine's graphics card (GPU), not a cloud voice API. Nothing about how you sound is sent anywhere to make it happen.

Silent reels in seconds

Don't want narration on a given reel? It can also render caption-only and silent, quickly, when that's the better fit for a post.

Publish

Scheduled and checked before anything goes out

Auto-publishing currently covers four platforms. The rest of the outputs (Instagram, slide carousels, quote cards, video, audio) are generated ready to post, for you to upload wherever that platform's own app or rules require.

Scheduling and auto-publish: Bluesky, X, Threads, LinkedIn

Set a time, and the studio posts for you on these four platforms, with automatic retries if a publish attempt fails and a Telegram alert if it can't recover.

Deterministic preflight checks

Before you approve a batch, the studio runs fixed, rule-based checks (not another AI guess) for things like posts that are over a platform's character limit, required fields left empty, or a reel script that won't fit its target length. Problems surface before publishing, not after.

You approve before it posts

Generated drafts sit in front of you first. Nothing publishes without you seeing and approving it.

Measure

What happened after it went out

Followers

Tracked over time so growth (or the lack of it) is visible, not guessed at.

Engagement per post

Each published post is tracked individually, so you can see which formats and topics actually land.

Words written

A running count of your own writing output, tracked alongside the distribution side.