Build a prompt library your editors can trust
Organize prompts around real editorial jobs, bounded evidence, reviewable outputs, and repeatable comparisons.
Explore AI assistance as an editorial activity with bounded source material, explicit review criteria, and human-owned release authority.
The human-review guide describes the full handoff from brief to accepted revision. The prompt-library guide focuses on making recurring assignments maintainable. Read them together when introducing AI into an existing publication, and keep a comparison set that includes missing evidence rather than only ideal inputs. A well-written draft is a candidate for review, not a substitute for it.
Choose the article closest to the decision you are making today. Note the source record, approval boundary, destination, and verification method it proposes. Compare those assumptions with the system you actually operate. Keep unresolved questions visible, and follow the related topic links when you need to examine a specific platform or implementation layer in more detail.
Organize prompts around real editorial jobs, bounded evidence, reviewable outputs, and repeatable comparisons.
Build a source-first review process that separates generated suggestions, editorial decisions, and the authority to publish.