# AI Security Labs Public Data

Curated public-safe outputs for the Labs product shell.

Source package: `AI Security Range`

The executable lab runtime remains upstream. This repo ships the public-facing product surface, curated scenario snapshots, metrics, charts, and roadmap content.

Generated files:
- `catalog.v1.json`
- `attack-range/overview.v1.json`
- `attack-range/metrics.v1.json`
- `attack-range/scenarios.v1.json`
- `attack-range/generation-media.v1.json`
- `trust-scanner/scorecard.v1.json`
- `trust-scanner/samples.v1.json`
- `trust-scanner/fixtures/real_targets/*.json`
- `trust-scanner/outputs/real_targets/*.report.json`
- `charts/chart_labs_range_generation_attacks_by_type_monthly.json`
- `charts/chart_labs_range_synthetic_media_attack_success_rate.json`
- `charts/chart_labs_range_output_safety_control_effectiveness.json`
- `charts/chart_labs_range_evidence_capture_completeness.json`

This directory is intentionally curated. It is not a mirror of the upstream runtime repo layout, and it should stay public-safe.