flaibo
AI Meta Prompts

Generate Prompt from Examples

Reverse-engineer a prompt from examples of our CEO's LinkedIn posts

No card required

What this brief specifies

Goal

The output needs to be a prompt an internal comms person can reuse to draft LinkedIn posts in the CEO's voice, drafts she'll edit but that should already feel substantially like her.

Examples Provided

The only raw material is five of her LinkedIn posts from the past six months, each 150 to 250 words, each opening on a counter-intuitive claim and closing on a soft question.

Observed Patterns

The patterns spotted so far: no hashtags, always one specific number or name, no humble-brag openings, and posts that often anchor to a specific moment from her week.

What makes this hard

Reverse-engineering a reusable prompt from only five posts means separating the CEO's repeatable structure from the specific topics those five happened to cover, and the two are easy to tangle together. The patterns span several layers at once: structural (counter-intuitive open, soft-question close), lexical (no hashtags, always one concrete number or name), tonal (no humble-brag openings), and habitual (anchoring to a real moment from her week), and naming all four risks reading as a rigid checklist rather than a voice. Get it too literal and future drafts overfit to whatever the five examples happened to be about; get it too abstract and the prompt loses exactly the specificity, one concrete number or name, that makes her posts sound like her instead of generic LinkedIn thought leadership.

FounderConsultantMarketingCreatorInfopreneurCorporate

Frequently asked questions

The next time a specific person's voice needs to be captured well enough for someone else to draft convincingly in it, the structural, lexical and tonal patterns that made this one recognizable do not have to be re-derived from a fresh batch of examples.

Ready to try this in Flaibo?

Get started free

Your context everywhere