skills/imago/references/factual_visuals_and_data_integrity.md

# Factual Visuals and Data Integrity

## Rule

A beautiful visual with invented facts is a failed output.

## Source hierarchy

Prefer:
1. user-provided authoritative data;
2. primary official sources;
3. strong secondary sources when primary material is unavailable;
4. clearly labeled assumptions only when the user explicitly permits them.

## Never invent

Do not fabricate:
- numeric values;
- percentages;
- dates;
- prices;
- counts;
- rankings;
- quotes;
- source names;
- labels;
- locations;
- factual comparisons.

## Derived values

When calculating:
- show or preserve the source values internally;
- use deterministic computation where tools are available;
- verify units and denominators;
- avoid mixing incompatible time periods or populations.

## Missing data

Choose one:
- omit it;
- mark it unavailable;
- use a clearly labeled placeholder for a template;
- ask one targeted question if the missing value is essential.

## Visual authority

Charts, dashboards, diagrams, maps, and infographics can look more authoritative than prose.

Do not visually imply precision the source does not support.

## Source conflicts

When sources disagree:
- prefer the authoritative/current source when the precedence is clear;
- otherwise surface the disagreement rather than picking a number silently.