bluue-infographic-series

Bluue Infographic Series

Convert information into a readable visual system, then generate one image or a coherent series of images. The primary job is information architecture: decide what deserves visual space, how many frames are needed, what visual form fits each idea, and which style can carry the whole series without sacrificing readability.

Treat the image as a communication surface, not a container for the entire source. If the content is too dense, split it into more frames instead of shrinking text or stacking more modules.

Defaults

Use these defaults unless the user overrides them:

• Output mode: generate images
• Aspect ratio: 9:16
• Language priority: explicit user request → dominant language of supplied content → English
• Long-form output: series by default
• Short or focused output: single image when one frame can carry the message clearly
• Series length: usually 3–7 frames; use fewer or more when the material clearly demands it
• Content type: infer one primary type from data, explainer, summary, process, general
• Art direction: select one style for the entire series and lock it before generating frame 1
• Palette: restrained; normally one base, one neutral/text system, and one accent family
• Image density: one primary message per frame
• Text density: sparse; split rather than cram
• Visual assets: prefer data-native and relationship-native forms such as charts, topology, timelines, thresholds, paths, and spatial hierarchy; use icons only when they communicate faster than the native structure
• User questions: ask only when uncertainty would materially change the output

Accept natural-language requests. Do not require the user to provide a parameter object.

Hard rules

1. One frame, one message. A frame may contain supporting evidence, but it must have one obvious takeaway.
2. One series, one art direction. Keep palette, typography logic, icon language, line weight, texture, chart styling, image treatment, and general visual grammar consistent across the full series.
3. Do not overload the canvas. If content does not fit at a readable size, create another frame.
4. Text fidelity is a release gate. Names, dates, percentages, units, labels, citations, and factual claims must match the resolved content brief.
5. Do not invent data. Use supplied or researched facts only. Clearly label illustrative values.
6. Charts must be truthful. Geometry, order, proportions, and units must correspond to the data.
7. Visual decoration stays subordinate to comprehension. Do not let texture, photography, collage, gradients, icons, or illustration compete with the message.
8. Do not default to stock-icon shorthand. Avoid icon-per-bullet layouts, rows of circular icon badges, and repeated avatar, building, document, shield, robot-head, AI-brain, checkmark, or warning-triangle symbols when the underlying system can be shown directly.
9. Default to 9:16. Use another ratio only when the user requests it or the target surface clearly requires it.
10. Preserve language intentionally. Do not translate supplied content unless the user asks or the output language is explicitly changed.
11. Generate frames separately. Never return a contact sheet or collage of the series unless the user explicitly asks for one.