Growth Autopilots
Your best post, reshaped for somewhere else. You still approve every variant before it publishes.
You wrote one good post. Turning it into a version that reads natively on three other platforms is an hour of work nobody enjoys. A Growth Autopilot does the reshaping and puts the results in your queue as drafts, so the review is a minute instead of an hour.
Drafts only until you approve them.
How it works
Auto-repurpose, step by step.
- 1
Pick a source
The platform you are already winning on. Its top posts are the raw material.
- 2
Pick a destination
The channel you want to grow. Rules for its format come from the registry.
- 3
Postonus adapts each post
Length, structure, and hashtag norms change to match the destination, not just the caption.
- 4
Review and go
Adapted posts arrive as drafts in your queue. Approve, edit, or bin them.
Capabilities
What you get with Auto-repurpose.
- Mines what already worked
- The source is your own top-performing posts, imported from a connected account, rather than a blank prompt. Repurposing beats generating because the idea is already validated.
- Reshapes to the format, not just the length
- Each destination has its own character budget, media rules, and conventions, read from the same registry the composer and the publish workers use. A 2,000-character LinkedIn post does not arrive on Bluesky truncated.
- Nothing publishes unreviewed
- Every adapted post lands as a draft. There is no setting that lets an autopilot publish straight to a channel, because that is how brand-damaging posts happen.
- Bounded, so it cannot run away
- Repurpose volume and how many autopilots can run at once are both capped, and the cap is reported in the error rather than hidden. A misconfigured rule cannot quietly spend your whole month of credits.
Why repurposing beats generating
A blank-page AI post is a guess. A post you already published, that already earned engagement, is evidence. Reusing the evidence is a much better bet than asking a model to invent something.
That is the whole design premise here. An autopilot never starts from nothing. It starts from a post of yours that worked, and its job is translation rather than invention.
What "adapted" actually means
Not truncation. The destination's real constraints drive the rewrite:
- Character budget. Bluesky counts graphemes against 300; a flag emoji costs 1 there and up to 11 if you count naively. LinkedIn gives you thousands. The same idea has to be a different length.
- Media rules. How many images a post accepts, and which types, differ per platform. A carousel that works on Instagram is not a valid Bluesky post.
- Convention. Hashtag density that reads normally on Instagram reads desperate on LinkedIn.
Those constraints come from one registry that the composer, the publish workers, and this feature all read. So the adapted post is valid where it is going — you are not discovering a length problem at publish time.
Repurpose guidance is per-platform data, not a prompt instruction. That is why a destination the platform rules say is a poor fit for a given source is simply not offered, instead of producing something odd.
The review step is not optional
There is no "publish automatically" switch. Every adapted post is a draft.
This is a deliberate product limit rather than a missing feature. An autopilot that could publish is one bad adaptation away from posting something embarrassing to a channel you cannot properly edit afterwards. Making approval explicit costs you a few seconds and removes that whole category of incident.
Approving is also what charges the post against your monthly per-platform quota, so your drafts folder can be as full as you like without spending anything.
Where the ceilings are
Two caps apply, and both name themselves when you hit them:
- How many posts a repurpose run may adapt.
- How many autopilots may run at the same time.
Both exist so a rule pointed at a large source library cannot spend a month of AI credits in one afternoon. If you hit one, the error tells you the limit rather than failing vaguely.
Found this useful? Pass it on.
Questions
Frequently asked questions Everything the first evaluation usually needs.
- Does the AI write my posts from scratch?
No. You write the post. An autopilot takes something you already published and re-fits it to another platform’s shape — character limits, structure, hashtag norms. You review every variant before it goes out.
- Where do the source posts come from?
From a connected account you nominate as the source. Postonus imports your existing posts from it, so the autopilot works from real performance rather than guesses.
- Can an autopilot publish without me?
No. Adapted posts arrive as drafts in your queue. Approving one is an explicit act, and that is also the point at which it starts counting against your monthly platform quota.
- What does this cost in AI credits?
A refine on a single platform costs 2 credits. A multi-platform refine costs the number of channels plus one, because it is one model call producing a tailored variant per channel. Your plan sets the monthly allowance.
- Which platforms can be a source or a destination?
The pairing is driven by per-platform repurpose guidance in the registry, so the composer only offers combinations that make sense. A short-form video platform is not offered as the source for a text-only channel.
- How is this different from resharing the same post everywhere?
Cross-posting sends identical text to every channel, which reads as automated. Repurposing changes the post to suit the destination. Postonus also does straightforward recycling if identical resharing is what you want — see evergreen recycling.
Keep reading
- Evergreen recyclingTag your timeless posts, attach a recycle rule, and they return to the queue on a cadence you set — with a per-post cooldown so the same post does not burn out.Read
- AI assistantWrite the post once, then refine it into a variant that fits each platform’s length and conventions. You review every version, and credits are metered upfront.Read
- Multi-platform publishingWrite once and publish to 16 platforms, with per-channel overrides where the platforms differ. Limits are enforced as you type, from the rules workers use.Read