Editorial standards

Publishing and review methodology

Promptgena is a prompt library and editing tool. This page explains what publication means, what it does not prove, and how readers can evaluate or challenge a listing.

Last updated: July 19, 2026 · Maintained by Can Kırkgöz

Publication workflow

Structured submission

A submission must include a title, substantive description, image, valid JSON object, supported-model labels, and the filters required by the builder. Public submissions enter a pending state.

Administrative decision

Pending submissions are not published automatically. An administrator approves or rejects them. Review records include the decision date, and rejection reasons can identify policy, quality, duplication, or completeness problems.

Usability checks

Publication review checks whether the JSON can be parsed, the prompt has a coherent purpose, filters target real fields, the example image and description correspond, and the listing gives a user enough context to adapt it.

Rights and safety checks

Uploaders remain responsible for rights and consent. Content that is deceptive, exploitative, illegal, sexually explicit, harmful, or incompatible with the community rules should not be published.

What “supported model” means

A compatibility label means that the prompt’s visual intent can reasonably be adapted to that model family. It is not a claim that identical output was reproduced in every service, version, interface, checkpoint, or account configuration. Reference handling, negative prompts, aspect ratios, safety systems, and parameter syntax vary.

Readers should use the model adaptation notes on each prompt and document the exact model version and settings used for important work. See the model adaptation guide for a controlled comparison workflow.

Prompt-page editorial layer

Every published prompt page turns the underlying JSON into a readable scene blueprint, visual treatment, composition summary, model-specific adaptation notes, quality checklist, and troubleshooting guidance. This makes the reasoning inspectable instead of presenting a large generated text block without context.

Extracted details come from the listing’s current JSON and metadata. General advice is clearly separated from prompt-specific fields. We do not invent a claim that a prompt was tested when no documented test record is available.

Corrections, duplicates, and removal

Listings may be revised, merged, unpublished, or removed when they are incomplete, substantially duplicated, misleading, unsafe, or subject to a valid rights request. The visible update date helps readers identify changed material. Users can report a concern through the contact address in the footer or the imprint and contact page.

Independence and limitations

Promptgena is independently operated and is not affiliated with or endorsed by OpenAI, Google, Midjourney, Black Forest Labs, Stability AI, or other model providers. Product names identify compatibility targets. Generated-by labels describe the tool reported for a sample image; they do not certify rights, quality, or reproducibility.

Model behavior changes. A prompt that works well today may require revision after a model update. Readers should treat the gallery as a documented starting point and make their own safety, rights, and output-quality assessment.