Human Attention Dashboard

Proof you can understand and control.

The Human Attention Engine can evaluate, store, read, report, and now interactively filter cognitive protection evidence.

🟢Interactive proof loop online

Operator guide

What am I looking at?

This dashboard shows whether a feature or interface protects human attention. Start with the summary, inspect the latest evaluation, use filters, then create a new proof row with the composer.

1Read the current proof state.
2Filter by green, yellow, or red decisions.
3Compose a new evaluation and store proof.
Total2
Green2
Yellow0
Red0
Average18.5/21

Latest evaluation

operator_composer_contract_20260430164541

🟢green

19/21

Created: 2026-04-30 07:45 UTC

Canonical flow

Evaluate → Store → Read → Report → Control

v4.5.0 adds interaction: refresh, filters, workspace input, and limit control.

Evaluation composer

Create proof from the dashboard.

🟢green19/21

Composer guide

How should I score?

Use 0 for harmful, 1 for weak, 2 for acceptable, and 3 for strong. The goal is not to flatter the system, but to protect the human mind.

0 — harmful1 — weak2 — acceptable3 — strong

Suggested presets

landing_page_clarity_reviewdashboard_cognitive_load_reviewai_assistant_response_reviewproduct_feature_attention_review

History

Evaluation timeline

DecisionItemTypeScoreCreated
🟢greenoperator_composer_contract_20260430164541feature19/212026-04-30 07:45 UTC
🟢greencalm_priority_panelfeature18/212026-04-30 04:42 UTC

Report preview

Proof report

# Human Attention Evaluation Report

## Summary

Total: 2

Green: 2 · Yellow: 0 · Red: 0

Average score: 18.5/21

## Timeline

🟢 operator_composer_contract_20260430164541 — 19/21 — 2026-04-30 07:45 UTC

🟢 calm_priority_panel — 18/21 — 2026-04-30 04:42 UTC