Editorial Standards

How we research, write, review, and maintain the guides on Promptic.

What we cover

Promptic publishes practical guides on prompt engineering and AI-agent development — the craft of getting reliable results from large language models and the systems built on top of them.

How we research

Every guide is grounded in current primary sources: model and API documentation from the labs, peer-reviewed and preprint research, and first-hand testing of the techniques we describe. We prefer primary documentation over secondary summaries, and we link to our sources so you can verify claims yourself.

Authorship and review

Articles carry a named author who stands behind the content. Where a guide has been substantively revised or re-verified, we show a "Reviewed" date so you can judge how current it is. A missing review date simply means the guide has not been formally re-reviewed since publication — not that it is out of date.

Corrections and updates

The field moves quickly, and so do we. When a model changes, a technique is superseded, or we get something wrong, we update the affected guide and record a new review date. If you spot an error, please contact us — we take corrections seriously.

Use of AI

We use AI tools to assist with drafting and research, but every guide is directed, edited, fact-checked, and approved by a human author who is accountable for what it says.