Editorial policy
Editorial Policy
How GPT Image Hub organizes, describes, reviews, and publishes prompt templates for human users and AI agents.
Prompt quality standards
Published prompts should be useful, specific, reusable, and clear enough for a person or AI agent to understand the intended output.
- Prompts should include subject, style, composition, and output constraints when relevant.
- Prompts should avoid vague filler when concrete camera, lighting, layout, or material details are better.
- Prompt records should include category, tags, model hints, and source fields when available.
Safety and rights
The library avoids presenting unsafe, deceptive, or rights-ambiguous content as recommended production guidance.
- Source URLs and authors are preserved when known.
- License fields are exposed per record when available.
- Public prompt pages should not hide materially different machine-only content from users.
Updates and corrections
Prompt records may be updated for clarity, metadata quality, translation quality, or source attribution.
- Dataset exports include update timestamps.
- The prompt dataset manifest describes available formats and distributions.
- Correction work prioritizes canonical prompt details and machine-readable metadata.